Top AI-Based Research Papers on Prior Art Search

Top AI Based Research Papers on Prior Art Search

Since searchable digital patent databases gained widespread acceptance, the IP industry has been greatly interested in moving beyond basic, keyword-based searching. Numerous startups and established companies have tried their hand at the problem of AI-powered patent searches. With each attempt, they have developed multiple techniques.

True AI search that goes beyond keyword matching to understand what you are looking for  has remained elusive for decades. Fortunately, a new wave of progress has risen in the last few years as a result of the foundational developments in natural language processing using deep neural networks.

Early developments were triggered by the popularization of word embedding techniques around 2014. This technique of word representation allows words with similar meanings to have a similar representation. Word embeddings are a precursor to a new era in patent searching. This technique on its own, however, is insufficient for true AI search.

From 2015 to 2017, contextualized word embeddings and various techniques for creating sentence embeddings were invented. Then, in 2018, transformer-based language models such as BERT significantly advanced things. BERT (Bidirectional Encoder Representations from Transformers) better understands the nuances and context of words in search queries and can match those queries with more relevant results.

Google also released a custom version of its BERT model trained on patent data. Parallel advancements in AI training, such as contrastive learning and learning-to-rank techniques, have enabled researchers to develop more robust and entirely new ways of searching patents.

Today is an exciting time in the history of patent search technology, with an improved AI appearing on the research landscape every couple of weeks. These advances will undoubtedly make it easier for people, both professional searchers and inventors, to explore and make sense of the patent data.

But what about you? What will make your research less time-consuming?

PQAI has taken the onus to constantly update our project, in which we are creating a one-stop online repository of all the research papers in the field of AI-based patent search.

Moreover, we’ll notify you when a new paper is published to keep you updated with the latest advancements in the industry. This way, you will have access to all the previous and upcoming information available for your research/development without having a flood of emails stuffing your inbox.

For your convenience, all the papers published since 2004 have been listed below (latest first) with  author names, the year published, and download links.

Title

A Two-Stage Deep Learning-Based System For Patent Citation Recommendation

Author(s)

Choi Jaewoong, Lee Jiho, Yoon Janghyeok, Jang Sion, Jaeyoung Kim, Sungchul Choi

Year of Publishing

2022

Published On

Springer

Affiliation

Pukyong National University, Netmarble AI Center, VUNO INC

PDF

Link 

Abstract

The increasing number of patents leads patent applicants and examiners to spend more time and cost on searching and citing prior patents. Deep learning has exhibited outstanding performance in the recommendation of movies, music, products, and paper citation. However, the application of deep learning in patent citation recommendation has not been addressed well. Despite many attempts to apply deep learning models to the patent domain, there is little attention to the patent citation recommendation. Since patent citation is determined according to a complex technological context beyond simply finding semantically similar preceding documents, it is necessary to understand the context in which the citation occurs. Therefore, we propose a dataset named as a PatentNet to capture technological citation context based on textual information, meta data and examiner citation information for about 110,000 patents. Also, this paper proposes a strong benchmark model considering the similarity of patent text as well as technological citation context using cooperative patent classification (CPC) code. The proposed model exploits a two-stage structure of selecting based on textual information and pre-trained CPC embedding values and re-ranking candidates using a trained deep learning model with examiner citation information. The proposed model achieved improved performance with an MRR of 0.2506 on the benchmarking dataset, outperforming the existing methods. The results obtained show that learning about the descriptive citation context, rather than simple text similarity, has an important influence on citation recommendation. The proposed model and dataset can help researchers to understand technological citation context and assist patent examiners or applicants to find prior patents to cite effectively.

Title

Multi-Document Summarization For Patent Documents Based On Generative Adversarial Network

Author(s)

Sunhye Kim, Byungun Yoon

Year of Publishing

2022

Published On

Science Direct

Affiliation

Dongguk University

PDF

Link 

Abstract

Given the exponential growth of patent documents, automatic patent summarization methods to facilitate the patent analysis process are in strong demand. Recently, the development of natural language processing (NLP), text-mining, and deep learning has greatly improved the performance of text summarization models for general documents. However, existing models cannot be successfully applied to patent documents, because patent documents describing an inventive technology and using domain-specific words have many differences from general documents. To address this challenge, we propose in this study a multi-patent summarization approach based on deep learning to generate an abstractive summarization considering the characteristics of a patent. Single patent summarization and multi-patent summarization were performed through a patent-specific feature extraction process, a summarization model based on generative adversarial network (GAN), and an inference process using topic modeling. The proposed model was verified by applying it to a patent in the drone technology field. In consequence, the proposed model performed better than existing deep learning summarization models. The proposed approach enables high-quality information summary for a large number of patent documents, which can be used by R&D researchers and decision-makers. In addition, it can provide a guideline for deep learning research using patent data.

Title

A Survey On Sentence Embedding Models Performance For Patent Analysis

Author(s)

Hamid Bekamiri, Daniel S. Hain, Roman Jurowetzki

Year of Publishing

2022

Published On

ArXiv

Affiliation

Cornell University

PDF

Link  

Abstract

Patent data is an important source of knowledge for innovation research, while the technological similarity between pairs of patents is a key enabling indicator for patent analysis. Recently researchers have been using patent vector space models based on different NLP embeddings models to calculate the technological similarity between pairs of patents to help better understand innovations, patent landscaping, technology mapping, and patent quality evaluation. More often than not, Text Embedding is a vital precursor to patent analysis tasks. A pertinent question then arises: How should we measure and evaluate the accuracy of these embeddings? To the best of our knowledge, there is no comprehensive survey that builds a clear delineation of embedding models’ performance for calculating patent similarity indicators. Therefore, in this study, we provide an overview of the accuracy of these algorithms based on patent classification performance and propose a standard library and dataset for assessing the accuracy of embeddings models based on PatentSBERTa approach. In a detailed discussion, we report the performance of the top 3 algorithms at section, class, and subclass levels. The results based on the first claim of patents show that PatentSBERTa, Bert-for-patents, and TF-IDF Weighted Word Embeddings have the best accuracy for computing sentence embeddings at the subclass level. According to the first results, the performance of the models in different classes varies, which shows researchers in patent analysis can utilize the results of this study to choose the best proper model based on the specific section of patent data they used.

Title

End To End Neural Retrieval For Patent Prior Art Search

Author(s)

Vasileios Stamatis

Year of Publishing

2022

Published On

Springer

Affiliation

International Hellenic University

PDF

Link 

Abstract

This research will examine neural retrieval methods for patent prior art search. One research direction is the federated search approach, where we proposed two new methods that solve the results merging problem in federated patent search using machine learning models. The methods are based on a centralized index containing samples of documents from all potential resources, and they implement machine learning models to predict comparable scores for the documents retrieved by different resources. The other research direction is the adaptation of end-to-end neural retrieval approaches to the patent characteristics such that the retrieval effectiveness will be increased. Off-the-self neural methods like BERT have lower effectiveness for patent prior art search. So, we adapt the BERT model to patent characteristics in order to increase retrieval performance. We propose a new gate-based document retrieval method and examine it in patent prior art search. The method combines a first-stage retrieval method using BM25 and a re-ranking approach where the BERT model is used as a gating function that operates on the BM25 score and modifies it according to the BERT relevance score. These experiments are based on two-stage retrieval approaches as neural models like BERT requires lots of computing power to be used. Eventually, the final part of the research will examine first-stage neural retrieval methods such as dense retrieval methods adapted to patent characteristics for prior art search.

Title

A Doc2vec And Local Outlier Factor Approach To Measuring The Novelty Of Patents

Author(s)

Daeseong Jeon, Joon MoAhn, Juram Kim, Changyong Lee

Year of Publishing

2022

Published On

Science Direct

Affiliation

Ulsan National Institute of Science and Technology, Korea University, Korea Institute of Science and Technology Information, Sogang University

PDF

Link 

Abstract

Patent analysis using text mining techniques is an effective way to identify novel technologies. However, the results of previous studies have been of limited use in practice because they require domain-specific knowledge and reflect the limited technological features of patents. As a remedy, this study proposes a machine learning approach to measuring the novelty of patents. At the heart of this approach are doc2vec to represent patents as vectors using textual information of patents and the local outlier factor to measure the novelty of patents on a numerical scale. A case study of 1,877 medical imaging technology patents confirms that our novelty scores are significantly correlated with the relevant patent indicators in the literature and that the novel patents identified have a higher technological impact on average. It is expected that the proposed approach could be useful as a complementary tool to support expert decision-making in identifying new technology opportunities, especially for small and medium-sized companies with limited technological knowledge and resources.

Title

Deep Learning For Patent Landscaping Using Transformer And Graph Embedding

Author(s)

Seokkyu Choi, Hyeonju Lee, Eunjeong Park, Sungchul Choi

Year of Publishing

2022

Published On

Science Direct

Affiliation

Gachon University, Industrial Application R&D Institute, Seoul National University, Seoul National University

PDF

Link 

Abstract

Patent landscaping is used to search for related patents during research and development projects. Patent landscaping is a crucial task required during the early stages of an R & D project to avoid the risk of patent infringement and to follow current trends in technology. The first task of patent landscaping is to extract the target patent for analysis from a patent database. Because patent classification for patent landscaping requires advanced human resources and can be tedious, the demand for automated patent classification has gradually increased. However, a shortage of well-defined benchmark datasets and comparable models makes it difficult to find related research studies. This paper proposes an automated patent classification model for patent landscaping based on transformer and graph embedding, both of which are drawn from deep learning. The proposed model uses a transformer architecture to derive text embedding from patent abstracts and uses a graph neural network to derive graph embedding from classification code co-occurrence information and concatenates them. Furthermore, we introduce four benchmark datasets to compare related research studies on patent landscaping. The obtained results showed prominent performance that was actually applicable to our dataset and comparable to the model using BERT, which has recently shown the best performance.

Title

Establish A Patent Risk Prediction Model For Emerging Technologies Using Deep Learning And Data Augmentation

Author(s)

Yung-Chang Chi, Hei-Chia Wang

Year of Publishing

2022

Published On

Science Direct

Affiliation

National Cheng Kung University, National Cheng Kung University

PDF

Link 

Abstract

Technology patents are considered the source and bedrock of emerging technologies. Patents create value in any enterprise. However, obtaining patents is time consuming, expensive, and risky; especially if the patent application is rejected. The development of new patents requires extensive costs and resources, but sometimes they may be similar to other patents once the technology is fully developed. They might lack relevant patentable features and as a result, fail to pass the patent examination, resulting in investment losses. Patent infringement is also an especially important topic for reducing the risk of legal damages of patent holders, applicants, and manufacturers. Patent examinations have so far been performed manually. Due to manpower and time limitations, the examination time is exceedingly long and inefficient. Current patent similarity comparison research, and the classification algorithms of text mining are most commonly employed to provide analyses of the possibility of examination approval, but there is insufficient discussion about the possibility of infringement. However, if a new technology or innovation can be accurately determined in advance whether it likely to pass or fail (and why), or is at risk of patent infringement, losses can be mitigated.

This research attempts to identify the issues involved in evaluating patent applications and infringement risks from existing patent databases. For each patent application, this research uses Convolutional Neural Networks, CNN + Long Short Term Memory Network, LSTM, prediction model, and the United States Patent and Trademark Office (USPTO) public utility patent application and reviews results based on keyword search. Then, data augmentation is utilized before performing model training; 10% of the approved and rejected applications are randomly selected as test cases, with the remaining 90% of the cases used to train the prediction model of this research in order to determine a model that can predict patent infringement and examination outcomes. Experimental results of the model in this study predicts that the accuracy of each classification is at least 87.7%, and can be used to find the classification of the reason for a rejection of a patent application failure.

Title

Pre-Trained Transformer-Based Classification For Automated Patentability Examination

Author(s)

Hao-Cheng Lo, Jung-Mei Chu

Year of Publishing

2021

Published On

IEEE

Affiliation

National Taiwan University

PDF

Link 

Abstract

Patentability examination, which means checking whether claims of a patent application meet the requirements for being patentable, is highly reliant on experts’ arduous endeavors entailing domain knowledge. Therefore, automated patentability examination would be the immediate priority, though under-appreciated. In this work, being the first to cast deep-learning light on automated patentability examination, we formulate this task as a multi-label text classification problem, which is challenging due to learning cross-sectional characteristics of abstract requirements (labels) from text content replete with inventive terms. To address this problem, we fine-tune downstream multi-label classification models over pre-trained transformer variants (BERT-BaseLarge, RoBERTa-BaseLarge, and XLNet) in light of their state-of-the-art achievements on many tasks. On a large USPTO patent database, we assess the performance of our models and find the model outperforming others based on the metrics, namely micro-precision, micro-recall, and micro-F1.

Title

PatentNet: Multi-Label Classification Of Patent Documents Using Deep Learning Based Language Understanding

Author(s)

Arousha Haghighian Roudsari, Jafar Afshar, Wookey Lee, Suan Lee 

Year of Publishing

2021

Published On

Springer

Affiliation

Inha University, Semyung University

PDF

Link 

Abstract

Patent classification is an expensive and time-consuming task that has conventionally been performed by domain experts. However, the increase in the number of filed patents and the complexity of the documents make the classification task challenging. The text used in patent documents is not always written in a way to efficiently convey knowledge. Moreover, patent classification is a multi-label classification task with a large number of labels, which makes the problem even more complicated. Hence, automating this expensive and laborious task is essential for assisting domain experts in managing patent documents, facilitating reliable search, retrieval, and further patent analysis tasks. Transfer learning and pre-trained language models have recently achieved state-of-the-art results in many Natural Language Processing tasks. In this work, we focus on investigating the effect of fine-tuning the pre-trained language models, namely, BERT, XLNet, RoBERTa, and ELECTRA, for the essential task of multi-label patent classification. We compare these models with the baseline deep-learning approaches used for patent classification. We use various word embeddings to enhance the performance of the baseline models. The publicly available USPTO-2M patent classification benchmark and M-patent datasets are used for conducting experiments. We conclude that fine-tuning the pre-trained language models on the patent text improves the multi-label patent classification performance. Our findings indicate that XLNet performs the best and achieves a new state-of-the-art classification performance with respect to precision, recall, F1 measure, as well as coverage error, and LRAP.

Title

PatentSBERTa: A Deep NLP Based Hybrid Model For Patent Distance And Classification Using Augmented SBERT

Author(s)

Hamid Bekamiri, Daniel S. Hain, Roman Jurowetzki

Year of Publishing

2021

Published On

ArXiv

Affiliation

Aalborg University Business School

PDF

Link 

Abstract

This study provides an efficient approach for using text data to calculate patent-to-patent (p2p) technological similarity, and presents a hybrid framework for leveraging the resulting p2p similarity for applications such as semantic search and automated patent classification. We create embeddings using Sentence-BERT (SBERT) based on patent claims. To further increase the patent embedding quality, we use transformer models based on SBERT and RoBERT, and apply the augmented approach for fine-tuning SBERT by in-domain supervised patent claims data. We leverage SBERTs efficiency in creating embedding distance measures to map p2p similarity in large sets of patent data. We deploy our framework for classification with a simple Nearest Neighbors (KNN) model that predicts Cooperative Patent Classification (CPC) of a patent based on the class assignment of the K patents with the highest p2p similarity. We thereby validate that the p2p similarity captures their technological features in terms of CPC overlap, and at the same demonstrate the usefulness of this approach for automatic patent classification based on text data. Furthermore, the presented classification framework is simple and the results easy to interpret and evaluate by end-users. In the out-of-sample model validation, we are able to perform a multi-label prediction of all assigned CPC classes on the subclass (663) level on 1,492,294 patents with an accuracy of 54% and F1 score > 66%, which suggests that our model outperforms the current state-of-the-art in text-based multi-label and multi-class patent classification. We furthermore discuss the applicability of the presented framework for semantic IP search, patent landscaping, and technology intelligence. We finally point towards a future research agenda for leveraging multi-source patent embeddings, their appropriateness across applications, as well as to improve and validate patent embeddings by creating domain-expert curated Semantic Textual Similarity (STS) benchmark datasets.

Title

A Multi-Task Approach To Neural Multi-Label Hierarchical Patent Classification Using Transformers

Author(s)

Subhash Chandra Pujari, Annemarie Friedrich, Jannik Strotgen ¨

Year of Publishing

2021

Published On

Github

Affiliation

Bosch Center for Artificial Intelligence, Heidelberg University

PDF

Link 

Abstract

With the aim of facilitating internal processes as well as search applications, patent offices categorize documents into taxonomies such as the Cooperative Patent Categorization. This task corresponds to a multi-label hierarchical

text classification problem. Recent approaches based on pre-trained neural language models have shown promising performance by focusing on leaf-level label

prediction. Prior works using intrinsically hierarchical algorithms, which learn a

separate classifier for each node in the hierarchy, have also demonstrated their effectiveness despite being based on symbolic feature inventories. However, training one transformer-based classifier per node is computationally infeasible due

to memory constraints. In this work, we propose a Transformer-based Multi-task

Model (TMM) overcoming this limitation. Using a multi-task setup and sharing

a single underlying language model, we train one classifier per node. To the best

of our knowledge, our work constitutes the first approach to patent classification combining transformers and hierarchical algorithms. We outperform several

non-neural and neural baselines on the WIPO-alpha dataset as well as on a new

dataset of 70k patents, which we publish along with this work. Our analysis reveals that our approach achieves much higher recall while keeping precision high.

Strong increases on macro-average scores demonstrate that our model also performs much better for infrequent labels. An extended version of the model with

additional connections reflecting the label taxonomy results in a further increase

of recall especially at the lower levels of the hierarchy.

Title

PQPS: Prior-Art Query-Based Patent Summarizer Using RBM And Bi-LSTM

Author(s)

Girthana Kumaravel, Swamynathan Sankaranarayanan

Year of Publishing

2021

Published On

Hindawi

Affiliation

Anna University

PDF

Link 

Abstract

A prior-art search on patents ascertains the patentability constraints of the invention through an organized review of prior-art document sources. This search technique poses challenges because of the inherent vocabulary mismatch problem. Manual processing of every retrieved relevant patent in its entirety is a tedious and time-consuming job that demands automated patent summarization for ease of access. This paper employs deep learning models for summarization as they take advantage of the massive dataset present in the patents to improve the summary coherence. This work presents a novel approach of patent summarization named PQPS: prior-art query-based patent summarizer using restricted Boltzmann machine (RBM) and bidirectional long short-term memory (Bi-LSTM) models. The PQPS also addresses the vocabulary mismatch problem through query expansion with knowledge bases such as domain ontology and WordNet. It further enhances the retrieval rate through topic modeling and bibliographic coupling of citations. The experiments analyze various interlinked smart device patent sample sets. The proposed PQPS demonstrates that retrievability increases both in extractive and abstractive summaries.

Title

Artificial Intelligence For Patent Prior Art Searching

Author(s)

Rossitza Setchi, Irena Spasić, Jeffrey Morgan, Christopher Harrison, Richard Corken

Year of Publishing

2021

Published On

Science Direct

Affiliation

Cardiff University, Intellectual Property Office UK

PDF

Link 

Abstract

This study explored how artificial intelligence (AI) could assist patent examiners as part of the prior art search process. The proof-of-concept allowed experimentation with different AI techniques to suggest search terms, retrieve most relevant documents, rank them and visualise their content. The study suggested that AI is less effective in formulating search queries but can reduce the time and cost of the process of sifting through a large number of patents. The study highlighted the importance of the humanin-the-loop approach and the need for better tools for human-centred decision and performance support in prior art searching.

Title

A Survey On Deep Learning For Patent Analysis

Author(s)

Ralf Krestel, Renukswamy Chikkamath, Christoph Hewel, Julian  Risch

Year of Publishing

2021

Published On

Research Gate

Affiliation

University of Potsdam, University of Passau, BETTEN & RESCH

PDF

Link 

Abstract

Patent document collections are an immense source of knowledge for research and innovation communities worldwide. The rapid growth of the number of patent documents poses an enormous challenge for retrieving and analyzing information from this source in an effective manner. Based on deep learning methods for natural language processing, novel approaches have been developed in the field of patent analysis. The goal of these approaches is to reduce costs by automating tasks that previously only domain experts could solve. In this article, we provide a comprehensive survey of the application of deep learning for patent analysis. We summarize the state-of-the-art techniques and describe how they are applied to various tasks in the patent domain. In a detailed discussion, we categorize 40 papers based on the dataset, the representation, and the deep learning architecture that were used, as well as the patent analysis task that was targeted. With our survey, we aim to foster future research at the intersection of patent analysis and deep learning and we conclude by listing promising paths for future work.

Title

Patent Sentiment Analysis To Highlight Patent Paragraphs

Author(s)

Renukswamy Chikkamath, Vishvapalsinhji Ramsinh Parmar, Christoph Hewel, Markus Endres

Year of Publishing

2021

Published On

ArXiv

Affiliation

University of Passau, BETTEN & RESCH

PDF

Link 

Abstract

Given a patent document, identifying distinct semantic annotations is an interesting research aspect. Text annotation helps the patent practitioners such as examiners and patent attorneys to quickly identify the key arguments of any invention, successively providing a timely marking of a patent text. In the process of manual patent analysis, to attain better readability, recognising the semantic information by marking paragraphs is in practice. This semantic annotation process is laborious and time-consuming. To alleviate such a problem, we proposed a novel dataset to train Machine Learning algorithms to automate the highlighting process. The contributions of this work are: i) we developed a multi-class, novel dataset of size 150k samples by traversing USPTO patents over a decade, ii) articulated statistics and distributions of data using imperative exploratory data analysis, iii) baseline Machine Learning models are developed to utilize the dataset to address patent paragraph highlighting task, iv) dataset and codes relating to this task are open-sourced through a dedicated GIT web page: https://github.com/Renuk9390/Patent_Sentiment_Analysis, and v) future path to extend this work using Deep Learning and domain specific pre-trained language models to develop a tool to highlight is provided. This work assist patent practitioners in highlighting semantic information automatically and aid to create a sustainable and efficient patent analysis using the aptitude of Machine Learning.

Title

Artificial Intelligence Technology Analysis Using Artificial Intelligence Patent Through Deep Learning Model And Vector Space Model

Author(s)

Yongmin Yoo, Dongjin Lim, Kyungsun Kim

Year of Publishing

2021

Published On

ArXiv

Affiliation

NHN Diquest

PDF

Link 

Abstract

Thanks to rapid development of artificial intelligence technology in recent years, the current artificial intelligence technology is contributing to many part of society. Education, environment, medical care, military, tourism, economy, politics, etc. are having a very large impact on society as a whole. For example, in the field of education, there is an artificial intelligence tutoring system that automatically assigns tutors based on student’s level. In the field of economics, there are quantitative investment methods that automatically analyze large amounts of data to find investment laws to create investment models or predict changes in financial markets. As such, artificial intelligence technology is being used in various fields. So, it is very important to know exactly what factors have an important influence on each field of artificial intelligence technology and how the relationship between each field is connected. Therefore, it is necessary to analyze artificial intelligence technology in each field. In this paper, we analyze patent documents related to artificial intelligence technology. We propose a method for keyword analysis within factors using artificial intelligence patent data sets for artificial intelligence technology analysis. This is a model that relies on feature engineering based on deep learning model named KeyBERT, and using vector space model. A case study of collecting and analyzing artificial intelligence patent data was conducted to show how the proposed model can be applied to real world problem

Title

Identifying Artificial Intelligence (AI) Invention: A Novel AI Patent Dataset

Author(s)

Alexander V. Giczy, Nicholas A. Pairolero, Andrew Toole

Year of Publishing

2021

Published On

SSRN

Affiliation

United States Patent and Trademark Office

PDF

Link 

Abstract

Artificial Intelligence (AI) is an area of increasing scholarly and policy interest. To help researchers, policymakers, and the public, this paper describes a novel dataset identifying AI in over 13.2 million patents and pre-grant publications (PGPubs). The dataset, called the Artificial Intelligence Patent Dataset (AIPD), was constructed using machine learning models for each of eight AI component technologies covering areas such as natural language processing, AI hardware, and machine learning. The AIPD contains two data files, one identifying the patents and PGPubs predicted to contain AI and a second file containing the patent documents used to train the machine learning classification models. We also present several evaluation metrics based on manual review by patent examiners with focused expertise in AI, and show that our machine learning approach achieves state-of-the-art performance across existing alternatives in the literature. We believe releasing this dataset will strengthen policy formulation, encourage additional empirical work, and provide researchers with a common base for building empirical knowledge on the determinants and impacts of AI invention.

Title

BERT Based Freedom To Operate Patent Analysis

Author(s)

Michael Freunek, André Bodmer

Year of Publishing

2021

Published On

ArXiv

Affiliation

University of Berne

PDF

Link 

Abstract

In this paper we present a method to apply BERT to freedom to operate patent analysis and patent searches. According to the method, BERT is fine-tuned by training patent descriptions to the independent claims. Each description represents an invention which is protected by the corresponding claims. Such a trained BERT could be able to identify or order freedom to operate relevant patents based on a short description of an invention or product. We tested the method by training BERT on the patent class G06T1/00 and applied the trained BERT on five inventions classified in G06T1/60, described via DOCDB abstracts. The DOCDB abstract are available on ESPACENET of the European Patent Office. 

Title

BERT Based Patent Novelty Search By Training Claims To Their Own Description

Author(s)

Michael Freunek, André Bodmer

Year of Publishing

2021

Published On

ArXiv

Affiliation

University of Berne

PDF

Link 

Abstract

In this paper we present a method to concatenate patent claims to their own description. By applying this method, BERT trains suitable descriptions for claims. Such a trained BERT (claim-to-descriptionBERT) could be able to identify novelty relevant descriptions for patents. In addition, we introduce a new scoring scheme, relevance scoring or novelty scoring, to process the output of BERT in a meaningful way. We tested the method on patent applications by training BERT on the first claims of patents and corresponding descriptions. BERT’s output has been processed according to the relevance score and the results compared with the cited X documents in the search reports. The test showed that BERT has scored some of the cited X documents as highly relevant.

Title

Prior Art Search And Reranking For Generated Patent Text 

Author(s)

Jieh-Sheng Lee, Jieh Hsiang

Year of Publishing

2021

Published On

ArXiv

Affiliation

National Taiwan University

PDF

Link  

Abstract

Generative models, such as GPT-2, have demonstrated impressive results recently. A fundamental question we would like to address is: where did the generated text come from? This work is our initial effort toward answering the question by using prior art search. The purpose of the prior art search is to find the most similar prior text in the training data of GPT-2. We take a reranking approach and apply it to the patent domain. Specifically, we pre-train GPT-2 models from scratch by using the patent data from the USPTO. The input for the prior art search is the patent text generated by the GPT-2 model. We also pre-trained BERT models from scratch for converting patent text to embeddings. The steps of reranking are: (1) search the most similar text in the training data of GPT-2 by taking a bag-of-words ranking approach (BM25), (2) convert the search results in text format to BERT embeddings, and (3) provide the final result by ranking the BERT embeddings based on their similarities with the patent text generated by GPT-2. The experiments in this work show that such reranking is better than ranking with embeddings alone. However, our mixed results also indicate that calculating the semantic similarities among long text spans is still challenging. To our knowledge, this work is the first to implement a reranking system to identify retrospectively the most similar inputs to a GPT model based on its output.

Title

An Empirical Study On Patent Novelty Detection: A Novel Approach Using Machine Learning And Natural Language Processing

Author(s)

Renukswamy Chikkamath; Markus Endres; Lavanya Bayyapu; Christoph Hewel

Year of Publishing

2021

Published On

IEEE

Affiliation

University of Passau, BETTEN und RESCH

PDF

Link 

Abstract

Patent, a form of intellectual property often be in the first place when it comes to securing an invention. The legal boundaries created then will become key stages of turning an invention into a commercial product. In recent years, the unprecedented growth of patent applications has induced a great challenge to patent examiners. Novelty detection is one major step considered before and after filing a patent application to assure claimed inventions are new and non-obvious. This itself is considered as a salient stage of prior art search by patent applicants, patent examiners, patent attorneys, patent agent professionals. Management in terms of critical analysis of such a large scale of documents has become a challenge since missing an optimal, effective, and efficient system. To this end, we come up with a novel experimental case study to foster highly recursive and interactive tasks. We developed and investigated more than 50 machine learning models on the considered dataset. The contributions of this work include: (1) outlined and anticipated the importance of novelty detection in the patent domain, (2) develop various baseline models for novelty detection, (3) utilize immense contributions of deep learning towards NLP to improve baseline models, (4) assess the performance of every model by using different word embeddings like word2vec, glove, fasttext, and domain-specific embeddings, (5) a novel application of NBSVM algorithm on our dataset, and considered as exceptionally good of our models. We articulated the fulfillment of models using training and validation curves to prove seemingly negligible overfit or no overfit, in the hope that effective automation in novelty detection helps in driving down the routine prior art search efforts.

Title

Identifying Valuable Patents: A Deep Learning Approach

Author(s)

Leonidas Aristodemou

Year of Publishing

2020

Published On

University of Cambridge Repository

Affiliation

University of Cambridge

PDF

Link 

Abstract

Big data is increasingly available in all areas of manufacturing, which presents value for enabling a competitive data-driven economy. Increased data availability presents an opportunity to introduce the next generation of innovative technologies. Firms invest in innovation and patents to increase, maintain and sustain competitive advantage. Consequently, the valuation of patents is a key determinant in economic growth since patents are an important innovation indicator. Given the surge in patenting throughout the world, the interest in the value of patents has grown significantly. Traditionally, studies on patent value have focused on limited data availability restricted to a specific technology area using methods such as regression, and mostly using numeric and binary categoric data types. We propose the definition for intellectual property intelligence (IPI) as the data science of analysing large amount of IP information, specifically patent data, with artificial intelligence (AI) methodologies to discover relationships and trends in the data for decision making. With the rise of AI and the ability to analyse larger datasets of patents, we develop an AI deep learning methodology for the valuation of patents. To do that, we build a large USPTO dataset consisting of all granted patents from 1976-2019: (i) we collect, clean, collate and pre-process all the data from the USPTO (and the OECD patent quality indicators database); (ii) we transform the data into numeric, categoric, and text features so that we are able to input them to the deep learning model. More specifically, we transform the text (abstract, claims, summary, title) into feature vectors using our developed Doc2Vec vector space model (VSM), that we assess using the t-distributed stochastic neighbour embedding (t-SNE) visualisation. The dataset is made publicly available for researchers to efficiently and effectively run fairly complex data analysis. We propose an AI deep learning methodology for the valuation of patents to identify valuable patents. Using our developed dataset, we build AI deep learning models, which are based on deep and wide feed-forward artificial neural networks (ANN), with dropout, L2 penalty and batch normalisation regularisation layers, to forecast the value of patents with 12 ex-post patent value output proxies. These include the grant_lag, generality, quality_index_4, and forward citations, generality_index and renewals in three time horizons (t4, t8, t12). We associate these patent value proxies to their respective patent value dimension (economic, strategic and technological). We forecast patent value using ex-ante patent value input determinants, for a wide range of technological areas (using the IPC classes), and time horizon domains (short term in t4, medium term in t8, and long term in t12). We evaluate all our models using a variety of strategies (out-of-time test, out-of-sample test, k-Fold and random split cross validation), and transparently report all metrics (accuracy, confusion matrix, F1-score, false negative rate, log loss, mean absolute error, precision, recall). Our models have higher accuracy and macro average F1-scores, with low values for the training and validation losses compared to prior art. With increasing prediction horizons, we observe an increase in the macro average F1-scores for several of the proxies. In addition, we find that the composite index that takes into consideration more than one value dimension, has the combined highest accuracy and macro average F1-score, relative to single value dimension patent proxies. Moreover, we find that firms seem to file widely at the short term time horizon and then focus their technological competencies to established opportunities. Patent owners seem to renew their patents in the fear of losing out. Our study has moved away from relatively small datasets, limited to specific technology field, and allowed for reproducibility in other fields. We can tailor models to different technology area, with different patent value proxies, with different time horizons. This study proposes an AI methodology, which is based on deep learning, using deep and wide feed forward artificial neural networks, to predict the value of patents, which has academic and industrial implications. We predict the value of patents with a variety of output proxies, including composite index proxies, for different technology areas (IPC classifications) and time horizons. Since we use all USPTO granted patents from 1976-2019 to train our models, we can apply this approach to patents in any technology field. Our approach enables researchers and industry professionals to value patents using a variety of patent value proxies, based on different value dimensions, tailored to specific technology areas. The proposed AI deep learning approach could effectively support expert decision making (technology, innovation and IP managers etc.) in their decision making by providing fast, low cost, data-driven intellectual property intelligence (IPI) from big patent data. Firms with limited resources, i.e. small-medium enterprises (SMEs) can choose representative proxies to forecast patent value estimates, saving resources. Consequently, the proposed approach could efficiently support experts in their patent value judgement, policy making in the government’s investments in technological sectors of the future to support the economy, and patent offices with the AI approaches to analyse efficiently and effectively big patent data. We anticipate this research would be interesting for future researchers to expand the emerging field of IPI research and the skills they will need to perform IPI data-driven research with a variety of data sources and AI deep learning ANN approaches.

Title

PatentMatch: A Dataset For Matching Patent Claims & Prior Art

Author(s)

Julian Risch, Nicolas Alder, Christoph Hewel, Ralf Krestel

Year of Publishing

2020

Published On

ArXiv

Affiliation

University of Potsdam, BETTEN & RESCH

PDF

Link 

Abstract

Patent examiners need to solve a complex information retrieval task when they assess the novelty and inventive step of claims made in a patent application. Given a claim, they search for prior art, which comprises all relevant publicly available information. This time-consuming task requires a deep understanding of the respective technical domain and the patent-domain-specific language. For these reasons, we address the computer-assisted search for prior art by creating a training dataset for supervised machine learning called PatentMatch. It contains pairs of claims from patent applications and semantically corresponding text passages of different degrees from cited patent documents. Each pair has been labeled by technically-skilled patent examiners from the European Patent Office. Accordingly, the label indicates the degree of semantic correspondence (matching), i.e., whether the text passage is prejudicial to the novelty of the claimed invention or not. Preliminary experiments using a baseline system show that PatentMatch can indeed be used for training a binary text pair classifier on this challenging information retrieval task. The dataset is available online: https://hpi.de/naumann/s/patentmatch

Title

Prior Art Search Using Multi-modal Embedding Of Patent Documents

Author(s)

Myungchul Kang, Suan Lee, Wookey Lee

Year of Publishing

2020

Published On

IEEE

Affiliation

Inha University

PDF

Link 

Abstract

Due to the limitations of the existing prior art search methods, a new patent search paradigm can be innovated by the concepts based on a precise patent document embedding, and a real-time feedback. These concepts can be achieved by the following ideas. The latest language model BERT can be incorporated with the description drawing embedding so that the explorable user interactive model can be adopted to the patent domain for “Building an artificial intelligent patent search system.” Therefore, these methodologies mainly with the help of deep learning can solve the traditional labor-intensive and time-consuming prior art search.

Title

Patent Prior Art Search Using Deep Learning Language Model

Author(s)

Dylan Myungchul Kang, Charles Cheolgi Lee, Suan Lee, Wookey Lee

Year of Publishing

2020

Published On

ACM

Affiliation

Inha University, VOICE AI Institute, 

PDF

Link 

Abstract

A patent is one of the essential indicators of new technologies and business processes, which becomes the main driving force of the companies and even the national competitiveness as well, that has recently been submitted and exploited in a large scale of quantities of information sources. Since the number of patent processing personnel, however, can hardly keep up with the increasing number of patents, and thus may have been worried about from deteriorating the quality of examinations. In this regard, the advancement of deep learning for the language processing capabilities has been developed significantly so that the prior art search by the deep learning models also can be accomplished for the labor-intensive and expensive patent document search tasks. The prior art search requires differentiation tasks, usually with the sheer volume of relevant documents; thus, the recall is much more important than the precision, which is the primary difference from the conventional search engines. This paper addressed a method to effectively handle the patent documents using BERT, one of the major deep learning-based language models. We proved through experiments that our model had outperformed the conventional approaches and the combinations of the key components with the recall value of up to ‘94.29%’ from the real patent dataset.

Title

Construction And Evaluation Of Gold Standards For Patent Classification—A Case Study On Quantum Computing

Author(s)

Steve Harris, Anthony Trippe, David Challis, Nigel Swycher

Year of Publishing

2020

Published On

Science Direct

Affiliation

Aistemos Ltd, Patinformatics LLC

PDF

Link 

Abstract

This article discusses options for evaluation of patent and/or patent family classification algorithms by means of “gold standards”. It covers the creation criteria, and desirable attributes of evaluation mechanisms, then proposes an example gold standard, and discusses the results of applying the evaluation mechanism against the proposed gold standard and an existing commercial implementation.

Title

Research On Classification And Similarity Of Patent Citation Based On Deep Learning 

Author(s)

Yonghe Lu, Xin Xiong, Weiting Zhang, Jiaxin Liu, Ruijie Zhao 

Year of Publishing

2020

Published On

Springer

Affiliation

Sun Yat-sen University

PDF

Link 

Abstract

This paper proposes a patent citation classification model based on deep learning, and collects the patent datasets in text analysis and communication area from Google patent database to evaluate the classification effect of the model. At the same time, considering the technical relevance between the examiners’ citations and the pending patent, this paper proposes a hypothesis to take the output value of the model as the technology similarity of two patents. The rationality of the hypothesis is verified from the perspective of machine statistics and manual spot check. The experimental results show that the model effect based on deep learning proposed in this paper is significantly better than the traditional text representation and classification method, while having higher robustness than the method combining Doc2vec and traditional classification technology. In addition, we compare between the proposed method based on deep learning and the traditional similarity method by a triple verification. It shows that the proposed method is more accurate in calculating technology similarity of patents. And the results of manual sampling show that it is reasonable to use the output value of the proposed model to represent the technology similarity of patents.

Title

Using AI To Analyze Patent Claim Indefiniteness 

Author(s)

Dean Alderucci, Kevin Ashley

Year of Publishing

2020

Published On

Indiana University Repository

Affiliation

Carnegie Mellon University, University of Pittsburgh

PDF

Link 

Abstract

We describe how to use artificial intelligence (AI) techniques to partially automate a type of legal analysis, determining whether a patent claim satisfies the definiteness requirement. Although fully automating such a high-level cognitive task is well beyond state-of-the-art AI, we show that AI can nevertheless assist the decision maker in making this determination. Specifically, the use of custom AI technology can aid the decision maker by (1) mining patent text to rapidly bring relevant information to the decision maker’s attention, and (2) suggesting simple inferences that can be drawn from that information.

We begin by summarizing the law related to patent claim indefiniteness. A summary of existing case law allows us to identify the types of information that can be relevant to the legal determination of indefiniteness. This in turn guides us in designing AI software that processes a patent’s text to extract information that can be relevant to the legal analysis of indefiniteness. Some types of relevant information include whether terms in a claim are defined in the patent, whether terms in a claim are not mentioned in the patent’s specification, whether the claim includes nonstandard terms coined by the drafter of the patent, whether the claim relies on vaguely-specified measurements, and whether the patent’s specification discloses structure corresponding to a means-plus-function limitation

The AI software rapidly processes a patent’s text and identifies information that is relevant to the legal analysis. The software then provides the human decision maker with this information as well as simple metrics and inferences, such as the percentage of claim terms that are defined explicitly or by example, and whether terms that are coined by the drafter should be defined or renamed. This can provide the user with insights about a patent much faster than if the user read the entirety of the patent to locate the same information unaided.

Moreover, the software can aggregate the various types of information to “score” a claim (e.g., from 0 to 100) based on its risk of being deemed indefinite. For example, a claim containing only defined terms and lacking any vague measurements would score much lower in terms of risk than a claim with terms that are not only undefined but do not even appear in the patent’s specification. Once each claim in a patent is assigned such an indefiniteness score, the patent itself can be given an overall indefiniteness score.

Scoring groups of patents in this manner has further advantages even if the scores are blunt measurements. AI software ranks a group of patents (e.g., all patents owned by a company) by indefiniteness scores. This allows a very large set of patents to be quickly searched for patents that have the highest, or lowest, indefiniteness score. The results of such a search could be, e.g., the patents to target for detailed review in litigation, post-grant proceedings, or licensing negotiations. Finally, we present some considerations for refining and augmenting the proposed methods for partially automating the indefiniteness analysis, and more broadly other types of legal analysis.

Title

Patent Document Clustering With Deep Embeddings

Author(s)

Jaeyoung Kim, Janghyeok Yoon, Eunjeong Park, Sungchul Choi  

Year of Publishing

2020

Published On

Springer

Affiliation

Gachon University, Konkuk University

PDF

Link 

Abstract

The analysis of scientific and technical documents is crucial in the process of establishing science and technology strategies. One popular method for such analysis is for field experts to manually classify each scientific or technical document into one of several predefined technical categories. However, not only is manual classification error-prone and expensive, but it also requires extended efforts to handle frequent data updates. In contrast, machine learning and text mining techniques enable cheaper and faster operations, and can alleviate the burden on human resources. In this paper, we propose a method for extracting embedded feature vectors by applying a neural embedding approach for text features in patent documents and automatically clustering the embedding features by utilizing a deep embedding clustering method.

Title

Optimizing Neural Networks For Patent Classification

Author(s)

Louay Abdelgawad, Peter Kluegl, Erdan Genc, Stefan Falkner, Frank Hutter

Year of Publishing

2020

Published On

Springer

Affiliation

Konkuk University, Albert-Ludwigs University of Freiburg, 

PDF

Link 

Abstract

A great number of patents is filed everyday to the patent offices worldwide. Each of these patents has to be labeled by domain experts with one or many of thousands of categories. This process is not only extremely expensive but also overwhelming for the experts, due to the considerable increase of filed patents over the years and the increasing complexity of the hierarchical categorization structure. Therefore, it is critical to automate the manual classification process using a classification model. In this paper, the automation of the task is carried out based on recent advances in deep learning for NLP and compared to customized approaches. Moreover, an extensive optimization analysis grants insights about hyperparameter importance. Our optimized convolutional neural network achieves a new state-of-the-art performance of 55.02% accuracy on the public Wipo-Alpha dataset.

Title

Engineering Knowledge Graph For Keyword Discovery In Patent Search 

Author(s)

Serhad Sarica, Binyang Song, En Low, Jianxi Luo

Year of Publishing

2019

Published On

Cambridge Univeristy Press

Affiliation

Singapore University of Technology and Design

PDF

Link 

Abstract

Patent retrieval and analytics have become common tasks in engineering design and innovation. Keyword-based search is the most common method and the core of integrative methods for patent retrieval. Often searchers intuitively choose keywords according to their knowledge on the search interest which may limit the coverage of the retrieval. Although one can identify additional keywords via reading patent texts from prior searches to refine the query terms heuristically, the process is tedious, time-consuming, and prone to human errors. In this paper, we propose a method to automate and augment the heuristic and iterative keyword discovery process. Specifically, we train a semantic engineering knowledge graph on the full patent database using natural language processing and semantic analysis, and use it as the basis to retrieve and rank the keywords contained in the retrieved patents. On this basis, searchers do not need to read patent texts but just select among the recommended keywords to expand their queries. The proposed method improves the completeness of the search keyword set and reduces the human effort for the same task.

Title

A Novelty Detection Patent Mining Approach For Analyzing Technological Opportunities

Author(s)

Juite Wang, Yi-Jing Chen

Year of Publishing

2019

Published On

Science Direct

Affiliation

National Chung Hsing University, 

PDF

Link 

Abstract

Early opportunity identification is critical for technology-based firms seeking to develop technology or product strategies for competitive advantage in the future. This research develops a patent mining approach based on the novelty detection statistical technique to identify unusual patents that may provide a fresh idea for potential opportunities. A natural language processing technique, latent semantic analysis, is applied to extract hidden relations between words in patent documents for alleviating the vocabulary mismatch problem and reducing the cumbersome efforts of keyword selection by experts. The angle-based outlier detection method, a novelty detection statistical technique, is used to determine outlier patents that are distinct from the majority of collected patent documents in a high-dimensional data space. Finally, visualization tools are developed to analyze the identified outlier patents for exploring potential technological opportunities. The developed methodology is applied in the telehealth industry and research findings can help telehealth firms formulate their technology strategies.

Title

TechNet: Technology Semantic Network Based On Pate

Author(s)

Serhad Sarica, Jianxi Luo, Kristin L. Wood

Year of Publishing

2019

Published On

ArXiv

Affiliation

Singapore University of Technology and Design

PDF

Link 

Abstract

The growing developments in general semantic networks, knowledge graphs and ontology databases have motivated us to build a large-scale comprehensive semantic network of technology-related data for engineering knowledge discovery, technology search and retrieval, and artificial intelligence for engineering design and innovation. Specially, we constructed a technology semantic network (TechNet) that covers the elemental concepts in all domains of technology and their semantic associations by mining the complete U.S. patent database from 1976. To derive the TechNet, natural language processing techniques were utilized to extract terms from massive patent texts and recent word embedding algorithms were employed to vectorize such terms and establish their semantic relationships. We report and evaluate the TechNet for retrieving terms and their pairwise relevance that is meaningful from a technology and engineering design perspective. The TechNet may serve as an infrastructure to support a wide range of applications, e.g., technical text summaries, search query predictions, relational knowledge discovery, and design ideation support, in the context of engineering and technology, and complement or enrich existing semantic databases. To enable such applications, the TechNet is made public via an online interface and APIs for public users to retrieve technologyrelated terms and their relevancies

Title

Improving Chemical Named Entity Recognition In Patents With Contextualized Word Embeddings

Author(s)

Zenan Zhai, Dat Quoc Nguyen, Saber Akhondi, Camilo Thorne, Christian Druckenbrodt, Trevor Cohn, Michelle Gregory, Karin Verspoor

Year of Publishing

2019

Published On

ACL Anthology

Affiliation

The University of Melbourne

PDF

Link 

Abstract

Chemical patents are an important resource for chemical information. However, few chemical Named Entity Recognition (NER) systems have been evaluated on patent documents, due in part to their structural and linguistic complexity. In this paper, we explore the NER performance of a BiLSTM-CRF model utilising pre-trained word embeddings, character-level word representations and contextualized ELMo word representations for chemical patents. We compare word embeddings pre-trained on biomedical and chemical patent corpora. The effect of tokenizers optimized for the chemical domain on NER performance in chemical patents is also explored. The results on two patent corpora show that contextualized word representations generated from ELMo substantially improve chemical NER performance w.r.t. the current state-of-the-art. We also show that domain-specific resources such as word embeddings trained on chemical patents and chemical-specific tokenizers, have a positive impact on NER performance.

Title

Automating The Search For A Patent’s Prior Art With A Full Text Similarity Search

Author(s)

Lea Helmers, Franziska Horn, Franziska Biegler, Tim Oppermann, Klaus-Robert Müller

Year of Publishing

2019

Published On

ArXiv

Affiliation

Technische Universität Berlin, Meinig & Partner, Korea University, Max-Planck-Institut für Informatik

PDF

Link 

Abstract

More than ever, technical inventions are the symbol of our society’s advance. Patents guarantee their creators protection against infringement. For an invention being patentable, its novelty and inventiveness have to be assessed. Therefore, a search for published work that describes similar inventions to a given patent application needs to be performed. Currently, this so-called search for prior art is executed with semi-automatically composed keyword queries, which is not only time consuming, but also prone to errors. In particular, errors may systematically arise by the fact that different keywords for the same technical concepts may exist across disciplines. In this paper, a novel approach is proposed, where the full text of a given patent application is compared to existing patents using machine learning and natural language processing techniques to automatically detect inventions that are similar to the one described in the submitted document. Various state-of-the-art approaches for feature extraction and document comparison are evaluated. In addition to that, the quality of the current search process is assessed based on ratings of a domain expert. The evaluation results show that our automated approach, besides accelerating the search process, also improves the search results for prior art with respect to their quality.

Title

Patent Classification By Fine-Tuning BERT Language Model

Author(s)

Jieh-Sheng Lee, Jieh Hsiang

Year of Publishing

2019

Published On

ArXiv

Affiliation

National Taiwan University

PDF

Link 

Abstract

In this work we focus on fine-tuning a pre-trained BERT model and applying it to patent classification. When applied to large datasets of over two millions patents, our approach outperforms the state of the art by an approach using CNN with word embeddings. In addition, we focus on patent claims without other parts in patent documents. Our contributions include: (1) a new state-of-the-art result based on pretrained BERT model and fine-tuning for patent classification, (2) a large dataset USPTO-3M at the CPC subclass level with SQL statements that can be used by future researchers, (3) showing that patent claims alone are sufficient for classification task, in contrast to conventional wisdom.

Title

Automatic Pre-Search: An overview

Author(s)

Dominique Andlauer

Year of Publishing

2018

Published On

Science Direct

Affiliation

European Patent Office

PDF

Link 

Abstract

This paper describes the evolution of the EPO’s search tools over time and their envisaged revolution towards supporting a more or even fully automated search process. Regardless of whether the goal of fully automated search is achieved completely or only partially, the chosen approach will in any case bring about major improvements for both the EPO’s examiners and the prior art search community at large.

Title

De-Noising Documents With A Novelty Detection Method Utilizing Class Vectors

Author(s)

Lee Younghoon, Cho Sungzoona, Choi Jinhaeb 

Year of Publishing

2018

Published On

IOS Press

Affiliation

Seoul National University

PDF

Link 

Abstract

The classification of customer-voice data is an important matter in real business since it is necessary for customer-voice data to be delivered to relevant departments and responsible individuals. Additionally, customer-voice data typically includes several novel words, such as typo’s, informal terms, or exceedingly general words to discriminate between categories of customer-voice data. Furthermore, noisy data often has a negative effect on the classification task. In this study, advanced novelty detection method is proposed to utilize class vector that possessed high cosine similarity with words to effectively discriminate between classes. The class vector is considered as the centroid or the mean of each word vector distribution as derived from the neural embedding model, and the novelty score of each word is calculated and novel words are effectively detected. Each novelty score is calculated by improvements of GMM and KMC methods utilizing a class vector. The experiments verify the propriety of the proposed method with qualitative observations, and the application of the proposed method with quantitative experiments verifies the representational effectiveness and classification performance of customer-voice data. The experiment results indicate that the performance of a classification of customer-voice data improved with the application of the newly proposed novelty detection method in this study.

Title

DeepPatent: Patent Classification With Convolutional Neural Networks And Word Embedding

Author(s)

Shaobo Li, Jie Hu, Yuxin Cui, Jianjun Hu 

Year of Publishing

2018

Published On

Springer

Affiliation

Guizhou University, University of South Carolina

PDF

Link 

Abstract

Patent classification is an essential task in patent information management and patent knowledge mining. However, this task is still largely done manually due to the unsatisfactory performance of current algorithms. Recently, deep learning methods such as convolutional neural networks (CNN) have led to great progress in image processing, voice recognition, and speech recognition, which has yet to be applied to patent classification. We proposed DeepPatent, a deep learning algorithm for patent classification based on CNN and word vector embedding. We evaluated the algorithm on the standard patent classification benchmark dataset CLEF-IP and compared it with other algorithms in the CLEF-IP competition. Experiments showed that DeepPatent with automatic feature extraction achieved a classification precision of 83.98%, which outperformed all the existing algorithms that used the same information for training. Its performance is better than the state-of-art patent classifier with a precision of 83.50%, whose performance is, however, based on 4000 characters from the description section and a lot of feature engineering while DeepPatent only used the title and abstract information. DeepPatent is further tested on USPTO-2M, a patent classification benchmark data set that we contributed with 2,000,147 records after data cleaning of 2,679,443 USA raw utility patent documents in 637 categories at the subclass level. Our algorithms achieved a precision of 73.88%.

Title

A Hierarchical Feature Extraction Model For Multi-Label Mechanical Patent Classification

Author(s)

Jie Hu, Shaobo Li, Jianjun Hu, Guanci Yang

Year of Publishing

2018

Published On

MDPI

Affiliation

Guizhou University, University of South Carolina

PDF

Link 

Abstract

Various studies have focused on feature extraction methods for automatic patent classification in recent years. However, most of these approaches are based on the knowledge from experts in related domains. Here we propose a hierarchical feature extraction model (HFEM) for multi-label mechanical patent classification, which is able to capture both local features of phrases as well as global and temporal semantics. First, a n-gram feature extractor based on convolutional neural networks (CNNs) is designed to extract salient local lexical-level features. Next, a long dependency feature extraction model based on the bidirectional long–short-term memory (BiLSTM) neural network model is proposed to capture sequential correlations from higher-level sequence representations. Then the HFEM algorithm and its hierarchical feature extraction architecture are detailed. We establish the training, validation and test datasets, containing 72,532, 18,133, and 2679 mechanical patent documents, respectively, and then check the performance of HFEMs. Finally, we compared the results of the proposed HFEM and three other single neural network models, namely CNN, long–short-term memory (LSTM), and BiLSTM. The experimental results indicate that our proposed HFEM outperforms the other compared models in both precision and recall.

Title

Automated Patent Landscaping

Author(s)

Aaron Abood, Dave Feltenberger   

Year of Publishing

2018

Published On

Springer

Affiliation

Google Inc

PDF

Link 

Abstract

Patent landscaping is the process of finding patents related to a particular topic. It is important for companies, investors, governments, and academics seeking to gauge innovation and assess risk. However, there is no broadly recognized best approach to landscaping. Frequently, patent landscaping is a bespoke human-driven process that relies heavily on complex queries over bibliographic patent databases. In this paper, we present Automated Patent Landscaping, an approach that jointly leverages human domain expertise, heuristics based on patent metadata, and machine learning to generate high-quality patent landscapes with minimal effort. In particular, this paper describes a flexible automated methodology to construct a patent landscape for a topic based on an initial seed set of patents. This approach takes human-selected seed patents that are representative of a topic, such as operating systems, and uses structure inherent in patent data such as references and class codes to “expand” the seed set to a set of “probably-related” patents and anti-seed “probably-unrelated” patents. The expanded set of patents is then pruned with a semi-supervised machine learning model trained on seed and anti-seed patents. This removes patents from the expanded set that are unrelated to the topic and ensures a comprehensive and accurate landscape.

Title

User Interface For Managing And Refining Related Patent Terms

Author(s)

Girish Showkatramani, Arthi Krishna, Ye Jin, Aaron Pepe, Naresh Nula, Greg Gabel

Year of Publishing

2018

Published On

Springer 

Affiliation

United States Patent and Trademark Office

PDF

Link 

Abstract

One of the crucial aspects of the patent examination process is assessing the patentability of an invention by performing extensive keyword-based searches to identify related existing inventions (or lack thereof). The expertise of identifying the most effective keywords is a critical skill and time-intensive step in the examination process. Recently, word embedding techniques have demonstrated value in identifying related words. In word embedding, the vector representation of an individual word is computed based on its context, and so words with similar meaning exhibit similar vector representation. Using a number of alternate data sources and word embedding techniques we are able to generate a variety of word embedding models. For example, we initially clustered patent data based on the different areas of interests such as Computer Architecture or Biology, and used this data to train Word2Vec and fastText models. Even though the generated word embedding models were reliable and scalable, none of the models by itself was sophisticated enough to match an experts choice of keywords.

In this study, we have developed a user interface that allows domain experts to quickly evaluate several word embedding models and curate a more sophisticated set of related patent terms by combining results from several models or in some cases even augmenting to them by hand. Our application thereby seeks to provide a functional and usable centralized interface towards searching and identifying related terms in the patent domain.

Title

Supervised Approaches To Assign Cooperative Patent Classification (CPC) Codes To Patents

Author(s)

Tung Tran, Ramakanth Kavuluru

Year of Publishing

2017

Published On

NETLAB

Affiliation

University of Kentucky

PDF

Link 

Abstract

This paper re-introduces the problem of patent classification with respect to the new Cooperative Patent Classification (CPC) system. CPC has replaced the U.S. Patent Classification (USPC) coding system as the official patent classification system in 2013. We frame patent classification as a multi-label text classification problem in which the prediction for a test document is a set of labels and success is measured based on the micro-F1 measure. We propose a supervised classification system that exploits the hierarchical taxonomy of CPC as well as the citation records of a test patent; we also propose various label ranking and cut-off (calibration) methods as part of the system pipeline. To evaluate the system, we conducted experiments on U.S. patents released in 2010 and 2011 for over 600 labels that correspond to the “subclasses” at the third level in the CPC hierarchy. The best variant of our model achieves ≈ 70% in micro-F1 score and the results are statistically significant. To the best of our knowledge, this is the first effort to reinitiate the automated patent classification task under the new CPC coding scheme. 

Title

Patents Images Retrieval And Convolutional Neural Network Training Dataset Quality Improvement

Author(s)

Alla Kravets, Nikita Lebedev, Maxim Legenchenko 

Year of Publishing

2017

Published On

Atlantis Press

Affiliation

Volgograd State Technical University

PDF

Link 

Abstract

The paper considers the problem of the analysis of patents’ figures for formalization of subjective opinions of the patent office experts that reviews applications for inventions. Drawings omission may indicate an incomplete description of the invention and entail the rejection of patent applications and other problems. Since patent images, even if one considers images of the same type, class, etc., are unique, different from each other. Nowadays for image processing are applied neural networks with different architectures. Neural network, Convolutional neural network, Siamese neural network were considered in the research. 4 libraries (Theano, TensorFlow, Caffe, and Keras) were studied. The main contributions of the paper are the new classification of patents’ imaged, training dataset formation and quality improvement approach, and the software implementation for CNN training.

Title

An Initial Study Of Anchor Selection In Patent Link Discovery

Author(s)

Dilesha Seneviratne, Shlomo Geva, Guido Zuccon, Andrew Trotman

Year of Publishing

2017

Published On

University of Otago

Affiliation

Queensland University of Technology, University of Otago

PDF

Link  

Abstract

Patents are a source of technical knowledge, but often difficult to understand. Technological solutions that would help understand the knowledge expressed in patents can assist the creation of new knowledge, and inventions. This paper explores anchor text selection for linking patents to external knowledge sources such as web pages and prior patents. While link discovery has been investigated in other domains, e.g., Wikipedia and the medical domain, the application of linking patents has received little attention and it presents some unique challenges as this paper shows. The paper contributes: (1) a test collection investigating the identification of anchor text (entities) in patent link discovery, (2) a user experiment studying the selection of anchors by users, and (3) an evaluation of four popular unsupervised keyword ranking methods (TFIDF, BM25, Keyphraseness, Termex) to identify potential anchors to link

Title

A Literature Review On Patent Information Retrieval Techniques

Author(s)

Alok Khode, Sagar Jambhorkar

Year of Publishing

2017

Published On

Indian Journal of Science and Technology

Affiliation

Symbiosis International University, National Defense Academy

PDF

Link 

Abstract

Patents are critical intellectual assets for any competitive business. They can prove to be a gold mine if retrieved, analyzed and utilized appropriately. Patentability search is an important step in the patent process and missing out any relevant patent may cause expensive legal consequences. As worldwide patent collection is growing rapidly, retrieval of this enormous knowledge source has become complex and exhaustive. This paper attempts to review the studies carried out in enhancing the relevance effectiveness of patent information retrieval. Method/Analysis: Literature review presents various research works that have been carried out to yield better results in patent retrieval task by refining existing information retrieval techniques or by using standard approaches at the various stages of the patent retrieval task. This work exclusively looks at literatures dealing with retrieval of patent text. Findings: Patent retrieval is not a completely solved research domain and general information retrieval approaches do not prove effective in this domain as patents are special documents posing various retrieval challenges. The review also highlights future research directions and will help researchers working in the domain of patent retrieval. Application/Improvement: Considering the various techniques and frameworks available and their limitations, there is a lot of scope in the field of patent retrieval techniques which makes room for further research to be taken up in this domain.

Title

Examiner Assisted Automated Patents Search

Author(s)

Arthi Krishna, Brian Feldman, Joseph Wolf, Greg Gabel, Scott Beliveau, Thomas Beach

Year of Publishing

2016

Published On

AAAI

Affiliation

United States Patent And Trademark Office

PDF

Link 

Abstract

One of the most crucial and knowledge-intenstive steps of patent examination is the determination of prior art- evidence that the idea claimed by a patent is already known. Automated prior art retrieval algorithms, if effective, can assist expert examiners by identifying literature that would otherwise take substantial research to uncover. Our approach is to build a patent search algorithm which functions as a cognitive assistant to the patent searcher. Contrary to the approach of treating the search algorithm as a black box, all componetnts of the seach algorithm are explained, and these components expose controls that can be adjusted by the user. This level of transparency and interactivity of the algorithm not only enables the experts to get the best use of the tool, but also is crucial in gaining the trust of the users. In this paper we discuss the engineering of the cognitive assistant search tool, referred to as Sigma, and the various interactions it affords the users. The tool is currently being piloted to patent examiners in the unit 2427.

Title

User Interface For Customizing Patents Search: An Exploratory Study

Author(s)

Arthi Krishna, Brian Feldman, Joseph Wolf, Greg Gabel, Scott Beliveau, Thomas Beach

Year of Publishing

2016

Published On

Springer

Affiliation

United States Patent and Trademark Office

PDF

Link 

Abstract

Prior art searching is a critical and knowledge-intensive step in the examination process of a patent application. Historically, the approach to automated prior art searching is to determine a few keywords from the patent application and, based on simple text frequency matching of these keywords, retrieve published applications and patents. Several emerging techniques show promise to increase the accuracy of automated searching, including analysis of: named entity extraction, explanations of how patents are classified, relationships between references cited by the examiner, weighing words found in some sections of the patent application differently than others, and lastly using the examiners’ domain knowledge such as synonyms. These techniques are explored in this study. Our approach is firstly, to design a user interface that leverages the above-mentioned processing techniques for the user and secondly, to provide visual cues that can guide examiner to fine tune search algorithms. The user interface displays a number of controls that affect the behavior of the underlying search algorithm—a tag cloud of the top keywords used to retrieve patents, sliders for weights on the different sections of a patent application (e.g., abstract, claims, title or specification), and a list of synonyms and stop-words. Users are provided with visual icons that give quick indication of the quality of the results, such as whether the results share a feature with the patent-at-issue, such as both citing to the same reference or having a common classification. This exploratory study shows results of seven variations of the search algorithm on a test corpus of 100500 patent documents.

Title

On Term Selection Techniques For Patent Prior Art Search

Author(s)

Mona Golestan Far

Year of Publishing

2016

Published On

Australian National University

Affiliation

The Australian National University

PDF

Link 

Abstract

A patent is a set of exclusive rights granted to an inventor to protect his invention for a limited period of time. Patent prior art search involves finding previously granted patents, scientific articles, product descriptions, or any other published work that may be relevant to a new patent application. Many well-known information retrieval (IR) techniques (e.g., typical query expansion methods), which are proven effective for ad hoc search, are unsuccessful for patent prior art search. In this thesis, we mainly investigate the reasons that generic IR techniques are not effective for prior art search on the CLEF-IP test collection. First, we analyse the errors caused due to data curation and experimental settings like applying International Patent Classification codes assigned to the patent topics to filter the search results. Then, we investigate the influence of term selection on retrieval performance on the CLEF-IP prior art test collection, starting with the description section of the reference patent and using language models (LM) and BM25 scoring functions. We find that an oracular relevance feedback system, which extracts terms from the judged relevant documents far outperforms the baseline (i.e., 0.11 vs. 0.48) and performs twice as well on mean average precision (MAP) as the best participant in CLEF-IP 2010 (i.e., 0.22 vs. 0.48). We find a very clear term selection value threshold for use when choosing terms. We also notice that most of the useful feedback terms are actually present in the original query and hypothesise that the baseline system can be substantially improved by removing negative query terms. We try four simple automated approaches to identify negative terms for query reduction but we are unable to improve on the baseline performance with any of them. However, we show that a simple, minimal feedback interactive approach, where terms are selected from only the first retrieved relevant document outperforms the best result from CLEF-IP 2010, suggesting the promise of interactive methods for term selection in patent prior art search.

Title

A Study Of Query Reformulation For Patent Prior Art Search With Partial Patent Applications

Author(s)

Mohamed Reda Bouadjenek, Scott Sanner, Gabriela Ferraro

Year of Publishing

2015

Published On

HAL 

Affiliation

University of Montpellier, ‡Oregon State University

PDF

Link 

Abstract

Patents are used by legal entities to legally protect their inventions and represent a multi-billion dollar industry of licensing and litigation. In 2014, 326,033 patent applications were approved in the US alone – a number that has doubled in the past 15 years and which makes prior art search a daunting, but necessary task in the patent application process. In this work, we seek to investigate the efficacy of prior art search strategies from the perspective of the inventor who wishes to assess the patentability of their ideas prior to writing a full application. While much of the literature inspired by the evaluation framework of the CLEF-IP competition has aimed to assist patent examiners in assessing prior art for complete patent applications, less of this work has focused on patent search with queries representing partial applications. In the (partial) patent search setting, a query is often much longer than in other standard IR tasks, e.g., the description section may contain hundreds or even thousands of words. While the length of such queries may suggest query reduction strategies to remove irrelevant terms, intentional obfuscation and general language used in patents suggests that it may help to expand queries with additionally relevant terms. To assess the trade-offs among all of these pre-application prior art search strategies, we comparatively evaluate a variety of partial application search and query reformulation methods. Among numerous findings, querying with a full description, perhaps in conjunction with generic (non-patent specific) query reduction methods, is recommended for best performance. However, we also find that querying with an abstract represents the best trade-off in terms of writing effort vs. retrieval efficacy (i.e., querying with the description sections only lead to marginal improvements) and that for such relatively short queries, generic query expansion methods help

Title

Novelty-Focused Patent Mapping For Technology Opportunity Analysis

Author(s)

Changyong Lee, Bokyoung Kang, Juneseuk Shin

Year of Publishing

2014

Published On

Science Direct

Affiliation

Ulsan National Institute of Science and Technology, Seoul National University, Sungkyunkwan University

PDF

Link 

Abstract

Patent maps are an effective means of discovering potential technology opportunities. However, this method has been of limited use in practice since defining and interpreting patent vacancies, as surrogates for potential technology opportunities, tend to be intuitive and ambiguous. As a remedy, we propose an approach to detecting novel patents based on systematic processes and quantitative outcomes. At the heart of the proposed approach is the text mining to extract the patterns of word usage and the local outlier factor to measure the degree of novelty in a numerical scale. The meanings of potential technology opportunities become more explicit by identifying novel patents rather than patent vacancies that are usually represented as a simple set of keywords. Finally, a novelty-focused patent identification map is developed to explore the implications on novel patents. A case study of the patents about thermal management technology of light emitting diode (LED) is exemplified. We believe the proposed approach could be employed in various research areas, serving as a starting point for developing more general models.

Title

A Survey Of Automated Hierarchical Classification Of Patents

Author(s)

Juan Carlos Gomez, Marie-Francine Moens

Year of Publishing

2014

Published On

KU Leuven

Affiliation

KU Leuven

PDF

Link 

Abstract

In this era of “big data”, hundreds or even thousands of patent applications arrive every day to patent offices around the world. One of the first tasks of the professional analysts in patent offices is to assign classification codes to those patents based on their content. Such classification codes are usually organized in hierarchical structures of concepts. Traditionally the classification task has been done manually by professional experts. However, given the large amount of documents, the patent professionals are becoming overwhelmed. If we add that the hierarchical structures of classification are very complex (containing thousands of categories), reliable, fast and scalable methods and algorithms are needed to help the experts in patent classification tasks. This chapter describes, analyzes and reviews systems that, based on the textual content of patents, automatically classify such patents into a hierarchy of categories. This chapter focuses specially in the patent classification task applied for the International Patent Classification (IPC) hierarchy. The IPC is the most used classification structure to organize patents, it is world-wide recognized, and several other structures use or are based on it to ensure office inter-operability.

Title

A Hybrid Patent Prior Art Retrieval Approach Using Claim Structure And Description

Author(s)

Fu-Ren Lin, Ke-Ren Chen, Szu-Yin Lin

Year of Publishing

2013

Published On

Springer

Affiliation

National Tsing Hua University, Chung Yuan Christian University

PDF

Link 

Abstract

In the highly competitive business environment, companies use patents as the intellectual asset to gain strategic competiveness. Patent prior art retrieval is a nontrivial task for invalidity and patentability search, which could help enterprises to plan their R&D strategies, patent portfolio, and avoid patent infringement issues in the future. This study adopts an efficient and effective hybrid patent prior art retrieval approach using claim structure and patent description to enhance prior art retrieval performance in terms of recall rate and exam the robustness through performing experiments in a large dataset. We obtained the best result by combining the information of claim structure and top 70 % sentences in description. We have achieved the competitive result in terms of raising the recall rate with the proposed hybrid approach, which also demonstrated the usefulness of including claim structure into patent prior art retrieval system.

Title

Identifying Technological Opportunities Using The Novelty Detection Technique: A Case Of Laser Technology In Semiconductor Manufacturing

Author(s)

Youngjung Geum, Jeonghwan Jeon, Hyeonju Seol

Year of Publishing

2013

Published On

Taylor & Francis Online

Affiliation

Seoul National University, Korea Air Force Academy

PDF

Link 

Abstract

While identification of technological opportunities has received considerable attention, previous studies have some weaknesses in terms of subjectivity when finding the opportunities in practical terms. This paper proposes a systematic framework to identify technological opportunities, focusing on objective evidences which are specific and practical to be used in a business environment. To do this, we used patents as a source and employed a novelty detection technique whose primary object is detecting the novel pattern. To begin with, the patents are collected from the United States Patent and Trademark Office (USPTO) database. These patents are then pre-processed into a structured keyword vector that can represent the characteristics of each patent. These keyword vectors are then used to analyse the new and emerging pattern, using the novelty detection technique. As the final step, the results are analysed to identify the technological opportunities. A case study on laser technology in lithography is presented to show the proposed framework.

Title

An Ontology-Based Automatic Semantic Annotation Approach For Patent Document Retrieval In Product Innovation Design

Author(s)

Feng Wang, Lan Fen Lin, Zhou Yang

Year of Publishing

2013

Published On

Scientific

Affiliation

Zhejiang University

PDF

Link 

Abstract

Patent retrieval plays a very important role in product innovation design. However, current patent retrieval approaches lack semantic comprehension and association, and usually cannot capture the implicit useful knowledge at a semantic level. In order to improve the traditional patent search, this paper proposes a novel ontology-based automatic semantic annotation approach based on the thorough analysis of patent documents, which combines both structure and content characteristics, and integrates multiple techniques from various aspects. Multilayer semantic model is established to realize unified semantic representation. The approach first utilizes template schemes to extract the structure information from patent documents, and then identifies semantics of entities and relations between entities from the content based on natural language processing techniques and domain knowledge, and at last employs a heuristic pattern learning method to abstract patent technical features. Case study is provided to show that our approach can acquire multi-level patent semantic knowledge from multiple perspectives, and discover semantic correlations between patent documents, which can further promote the accurate patent semantic retrieval effectively.

Title

Recommending Patents Based On Latent Topics

Author(s)

Ralf Krestel, Padhraic Smyth

Year of Publishing

2013

Published On

Research Gate

Affiliation

Leibniz Information Centre for Economics

PDF

Link 

Abstract

The availability of large volumes of granted patents and applications, all publicly available on the Web, enables the use of sophisticated text mining and information retrieval methods to facilitate access and analysis of patents. In this paper we investigate techniques to automatically recommend patents given a query patent. This task is critical for a variety of patent-related analysis problems such as finding relevant citations, research of relevant prior art, and infringement analysis. We investigate the use of latent Dirichlet allocation and Dirichlet multinomial regression to represent patent documents and to compute similarity scores. We compare our methods with state-of-the-art document representations and retrieval techniques and demonstrate the effectiveness of our approach on a collection of US patent publications.

Title

A New Instrument For Technology Monitoring: Novelty In Patents Measured By Semantic Patent Analysis

Author(s)

Jan M. Gerken, Martin G. Moehrle

Year of Publishing

2012

Published On

Springer

Affiliation

University of Bremen

PDF

Link 

Abstract

Given that in terms of technology novel inventions are crucial factors for companies; this article contributes to the identification of inventions of high novelty in patent data. As companies are confronted with an information overflow, and having patents reviewed by experts is a time-consuming task, we introduce a new approach to the identification of inventions of high novelty: a specific form of semantic patent analysis. Subsequent to the introduction of the concept of novelty in patents, the classical method of semantic patent analysis will be adapted to support novelty measurement. By means of a case study from the automotive industry, we corroborate that semantic patent analysis is able to outperform available methods for the identification of inventions of high novelty. Accordingly, semantic patent information possesses the potential to enhance technology monitoring while reducing both costs and uncertainty in the identification of inventions of high novelty.

Title

Automatic IPC Encoding And Novelty Tracking For Effective

patent mining

Author(s)

Douglas Teodoro, Emilie Pasche, Dina Vishnyakova, Christian Lovis, Julien Gobeill, Patrick Ruch

Year of Publishing

2011

Published On

BiTeM

Affiliation

University of Geneva, University of Applied Sciences

PDF

Link 

Abstract

Accurate classification of patent documents according to the IPC system is vital for the interoperability between different patent offices and for the prior art search task involved in a patent application procedure. It is essential for companies and governments to track changes in technology in order to asses their investments and create new branches of novel solutions. In this paper, we present our experiments from the NTCIR-8 challenge to automate paper abstract classification into the IPC taxonomy and to create a technical trend map from it. We apply the k-NN algorithm in the classification process and manipulate the rank of the nearest neighbours to enhance our results. The technical trend map is created by detecting technologies and their effects passages in paper and patent abstracts. A CRF-based system enriched with handcrafted rules is used to detect technology, effect, attribute and value phrases in the abstracts. Our experiments use multi patent databases for training the system and paper abstracts as well as patent applications for testing purposes, thus characterising a cross database and cross genre task. In the subtask of Research Papers Classification, we achieve a MAP of 0.68, 0.50 and 0.30 for the English and 0.71, 0.50 and 0.30 for the J2E subclass, main group and subgroup classifiers respectively. In the Technical Trend Map Creation subtask, we achieve an F-score of 0.138 when detecting technology/effect elements in patent abstracts and 0.141 in paper abstracts. Our methodology provides competitive results for the state of the art, with the majority of our official runs being ranked within the top two for both trend map (papers) and IPC coding. That said we see room for improvements especially in the detection of technologies and attributes elements in abstracts. Finally, we believe that the subtask of Technical Trend Map Creation needs to be adjusted in order to better produce a patent map. The classification system is available online at http://pingu.unige.ch:8080/IPCCat.

Title

A KNN Research Paper Classification Method Based On Shared Nearest Neighbor 

Author(s)

Yun-lei Cai, Duo Ji ,Dong-feng Cai

Year of Publishing

2010

Published On

NII Japan

Affiliation

Shenyang Institute of Aeronautical Engineering,

PDF

Link 

Abstract

The patents cover almost all the latest, the most active innovative technical information in technical fields, therefore patent classification has great application value in the patent research domain. This paper presents a KNN text categorization method based on shared nearest neighbor, effectively combining the BM25 similarity calculation method and the Neighborhood Information of samples. The effectiveness of this method has been fully verified in the NTCIR-8 Patent Classification evaluation.

Title

Exploring Contextual Models In Chemical Patent Search

Author(s)

Jay Urbain, Ophir Frieder

Year of Publishing

2010

Published On

Research Gate

Affiliation

Milwaukee School of Engineering, Georgetown University

PDF

Link 

Abstract

We explore the development of probabilistic retrieval models for integrating term statistics with entity search using multiple levels of document context to improve the performance of chemical patent search. A distributed indexing model was developed to enable efficient named entity search and aggregation of term statistics at multiple levels of patent structure including individual words, sentences, claims, descriptions, abstracts, and titles. The system can be scaled to an arbitrary number of compute instances in a cloud computing environment to support concurrent indexing and query processing operations on large patent collections. The query processing algorithm for patent prior art search uses information extraction techniques to identify candidate entities and distinctive terms from the query patent’s title, abstract, description, and claim sections. Structured queries integrating terms and entities in context are automatically generated to test the validity of each section of potentially relevant patents. The system was deployed across 15 Amazon Web Services (AWS) Elastic Cloud Compute (EC2) instances to support efficient indexing and query processing of the relatively large 100G+ collection of chemical patent documents. We evaluated several retrieval models for integrating statistics of candidate entities with term statistics at multiple levels of patent structure to identify relevant patents for prior art search. Our top performing retrieval model integrating contextual evidence from multiple levels of patent structure resulted in bpref measurements of 0.8929 for the prior art search task, exceeding the top results reported from the 2009 TREC Chemistry track.

Title

Improving Retrievability Of Patents In Prior-Art Search

Author(s)

Shariq Bashir, Andreas Rauber

Year of Publishing

2010

Published On

Vienna University of Technology

Affiliation

Vienna University of Technology

PDF

Link 

Abstract

Prior-art search is an important task in patent retrieval. The success of this task relies upon the selection of relevant search queries. Typically terms for prior-art queries are extracted from the claim fields of query patents. However, due to the complex technical structure of patents, and presence of terms mismatch and vague terms, selecting relevant terms for queries is a difficult task. During evaluating the patents retrievability coverage of prior-art queries generated from query patents, a large bias toward a subset of the collection is experienced. A large number of patents either have a very low retrievability score or can not be discovered via any query. To increase the retrievability of patents, in this paper we expand prior-art queries generated from query patents using query expansion with pseudo relevance feedback. Missing terms from query patents are discovered from feedback patents, and better patents for relevance feedback are identified using a novel approach for checking their similarity with query patents. We specifically focus on how to automatically select better terms from query patents based on their proximity distribution with prior-art queries that are used as features for computing similarity. Our results show, that the coverage of prior-art queries can be increased significantly by incorporating relevant queries terms using query expansion.

Title

Experiments On Patent Retrieval At NTCIR-4 Workshop

Author(s)

Hironori Takeuchi, Naohiko Uramoto, Koichi Takeda

Year of Publishing

2004

Published On

NII Japan

Affiliation

Tokyo Research Laboratory, National Institute of Informatics 

PDF

Link

Abstract

In the Patent Retrieval Task in NTCIR-4 Workshop, the search topic is the claim in a patent document, so we use the claim text and the IPC information for the similarity calculations between the search topic and each patent document in the collection. We examined the effectiveness of the similarity measure between IPCs and the term weighting for the occurrence positions of the keyword attributes in the search topic. As a result, it was found that the search results are slightly improved by considering not just the text in the search topic but also the hierarchical structural information of the IPCs. In contrast, the term frequencies for the occurrence position of the attribute did not improve the retrieval result.

The above list covers the published papers in the field of AI-based patent search. We hope that it assists you in your ongoing research in creating such AI-based patent search engines. Bookmark this page to have the list handy for future reference.

We at PQAI (Patent Quality Artificial Intelligence) are working to create an open-source AI-based library of patent tools to accelerate innovation and improve patent quality. One of our efforts is PQAI’s AI-based prior-art search tool, a collaborative initiative that drives diversity and inclusion by creating a level-playing field for all researchers in terms of prior-art searches. Currently, only big corporations and patent offices have adequate resources for these exhaustive searches. However, PQAI is democratizing the process by allowing zero-budget prior-art checks. We are a non-profit organization and firmly believe in transparency and user privacy. We do not store your data or search queries on our servers unless you specifically ask to do so for future reference.

PQAI is always looking for talented minds to help with our initiative. Get involved with us if you want to collaborate, have questions, or just want to say hi.

If you are a researcher working with patent data, you can also check out PQAI’s Researcher Page. You can use our open-source libraries, AI models and datasets to accelerate your work. You are also always invited to contribute to PQAI to help the research community in the field of AI-based patent search.

How AI Can Help Identify Companies for Patent Licensing or Synergistic Collaboration

Using AI to Help Identify Companies for Patent Licensing

With the increase of filed patents worldwide, locating meaningful information in this ever-growing dataset is getting more complex and time-consuming. 

Artificial intelligence (AI) plays a critical role in simplifying this process. From providing better patent classification performance to retrieving more relevant documents, AI offers great accuracy and exceptional quality in less time with less human effort.

AI can also help bring a patent idea to market by identifying patent licensing and synergistic collaboration opportunities.

How To Identify Companies for Patent Licensing or Synergistic Collaboration

Monetizing a patent can be challenging. An alternative to commercializing your invention is entering a licensing agreement with an established company. This reduces your investment and saves you the trouble of producing, marketing, or selling your product or technology.

Patent licensing and collaboration have become significant sources of revenue for patent holders. IBM, Microsoft, Ericsson, and others generate over $1 billion every year in licensing revenues. At its peak, Qualcomm earned over six billion dollars in licensing revenue. 

Options for finding partners include:

Identify Companies with Inventions that Rely on Your Invention

These companies may need to license a patent in order to refine their own inventions. 

For example, if you have a patent on an improved battery, electrical vehicle companies may need your invention to enhance the performance of their vehicles.

Seek Out Companies With Inventions You Rely On

If an enterprise has patents on the inputs you need for your invention, you may be able to form a relationship to exploit both sets of patents. 

You could, for example, enter into a:

  • License
  • Cross license
  • Joint venture
  • Vendor agreement

This relationship could help both of you dominate your market through cooperation rather than competition.

Find Companies with Competing Products

Companies with products that compete with your patented invention might have an interest in licensing or purchasing your patent. Bear in mind however, that a competitor’s primary interest in your patent might be to prevent products based on that patent from reaching the market.

In the above cases, you’re seeking companies that  want  to patent concepts similar to what you have already patented

In some cases your patents may have even been used to reject these companies’ patent applications. This might increase the odds of a relationship because these companies might already be familiar with your patent.

The best path for exploiting your invention will depend on many factors including:

  • The current market
  • Stakeholders in the market
  • The existence or absence of competitive or alternative products
  • The size of the entities already in the market
  • The size of entities looking to enter the market

Based on these factors, you should be able to identify some potential licensees or partners.

How AI Can Play a Role in Finding Potential Licensees, Partners, and Opportunities

AI technology can be a huge asset when it comes to finding patent licensees and collaborative partners. It does this by searching for subject matter related to your patented inventions. It can then research the attributes of the applicants. For example, it can identify startups and distinguish between competitors and non-competitors.

PQAI and Its Solution for Finding Licensing and Collaboration Opportunities

PQAI is an open-source library of tools that use AI to search and evaluate patent documents. These tools are often used before filing a patent application to conduct prior art searches and assess the likelihood of obtaining a patent.

But PQAI’s powerful AI engine can also identify potential licensees, partners, and opportunities by identifying companies with similar technology. PQAI can do this because it can find similar concepts even if the documents do not use the exact language used in the search query. It can also process many documents and identify which have the greatest relevance.

How to Use PQAI to Search for Companies With Patents in Similar Technology 

Step 1: Go to the PQAI search engine. Add the abstract of the patent to be licensed into the search query. You can also enter some keywords or any search query in plain English in accordance with your patent.

Search Form PQAI

Step 2: Click on the “Insights” Tab.

Insights Tab PQAI

Step 3: The Insights tab will produce a list of companies filing patents in similar areas. The list of companies that PQAI presents is not exhaustive. It’s merely a representation of what’s possible. We need to train the AI engine to provide more helpful results.

Proving the Concept Using Real-Life Situations

We conducted three case studies using real-life situations to show the power of the PQAI engine when looking for licensees and partners. Kindly bear in mind that these case studies were conducted using the back end of PQAI (roughly 2,000 search results) and not the front end interface (roughly the top 100 search results). 

Case Study 1 – Patent Licensing Opportunities for Canon

When looking for a partner, you must find enterprises interested in the same or similar technology as your patent.

In  2013, Canon invented a method for improving the emission efficiency and lifetime of OLEDs by using a layer of organic aromatic hydrocarbon compounds. In 2019 Canon secured a European patent on this technology, EP2939288B1.

We used PQAI to identify companies that filed patents on similar technologies by searching for the abstract from Canon’s patent. The search results returned Samsung, LG, NEC, Panasonic, SEL, Tianma, Sumitomo Chemical, Japan Display, Universal Display, Idemitsu, Foxconn, and Toray.

PQAI correctly identified potential partners. As it turned out, one-quarter of the companies returned by PQAI had licensed the patent from Canon. In fact, Samsung eventually bought the patent.

Additionally, when we studied the patents of Canon’s licensees, we found that these companies have patents in similar technology as Canon.

Case Study 2 – Synergistic Collaboration With StartUps for Nestle

Sometimes, a startup company has a disruptive solution that can benefit an established company.

In 2018, Nestle was researching dietary supplements with collagen. They invented a powder  containing buffer salt and amino acid. This powder helped prevent the protein from denaturing or coagulating when the dietary supplement was combined with a protein like collagen. Nestle filed a European patent application for its invention, EP3873236A2.

We used PQAI to search for companies filing patents for technology similar to Nestle’s application. We filtered the search to identify patents from start-ups by restricting the search results to small and medium companies that had filed for the first time within the preceding five years. The search results returned three companies: Vicenna, Vital Proteins, and iSatori.

Once again, PQAI correctly identified potential partners. Nestle acquired Vital Proteins and after the acquisition, it filed two additional patent applications for collagen-based dietary supplements, WO2021204643A1 and WO2021204644A1.

Case Study 3 – Collaboration With Players Across Supply Chain

Companies across a supply chain sometimes seek solutions to the same industry problem, and they often innovate in the same stage of the supply chain.

Nestle developed a process for manufacturing milk powder by spray drying fresh milk followed by a homogenization process before evaporation. This process solved the problem of high viscosity during production. The company secured a patent on this process, EP1259117B1.

We used PQAI to search for companies that were working on similar technologies. The search results included several of Nestle’s competitors—Kraft, Nabisco, Danone, and Unilever. But it also returned two businesses in the supply chain—packaging company Tetra Laval and chemical company Stauffer Chemical.

Tetra Laval, in particular, appeared to be a good potential partner for Nestle:

  • Tetra Laval filed a patent application EP3311672A1 for a method of producing milk powder that reduces the viscosity and increases the high total solids content
  • Tetra Laval is not a competitor
  • Tetra Laval’s subsidiary Tetra Pak has an existing relationship with Nestle

While Tetra Laval has not developed a relationship with Nestle for producing milk powder, all the ingredients for a good partnership exist.

An Opportunity for You – PQAI is Open Source!

Identifying companies for licensing or partnerships can be  arduous, research-intensive and time-consuming. Potential collaborators can leverage PQAI to create advanced tools that find excellent patent licensing and synergistic collaboration opportunities with greater speed and accuracy

PQAI has laid the foundation for potent AI-powered patent tools. As a non-profit open-source initiative, its mission is to accelerate innovation and improve patent search quality. PQAI is dedicated to establishing an open-source forum of IP tools that will drive critical changes in the IP ecosystem in a manner similar to what Linux did in computing.

To learn more about PQAI, visit PQAI’s website. And to become a collaborator in this open source project, review PQAI’s GitHub Directory.

5 Free and Possibly More Efficient Alternatives to Google Patents

Alternative to Google Patents

A Google Patents search requires the searcher to create complex keyword strings and sift through thousands of documents to find relevant prior art. 

In addition, some users are skeptical about using Google products due to a lack of trust and privacy issues. Although Google boasts strong security features, some doubts can remain: will my idea stay confidential if I run it through the Google Patents search portal

If you’re looking for an alternative to the Google Patents search portal for any of these reasons, you’ve hit the bull’s eye.

In this article, we’ll walk you through five free alternatives to Google Patents :

  1. PQAI—Patent Quality through Artificial Intelligence
  2. Lens 
  3. PATENTSCOPE 
  4. USPTO
  5. Espacenet

PQAI – Patent Quality through Artificial Intelligence

PQAI is an open-source library of patent tools. It offers a prior art search engine that accepts the invention query in plain English, making it simple for inventors to use. In addition, PQAI is AI-powered—it returns limited yet most relevant results. 

With PQAI, users can search a vast data set of ~11 million US patents and applications & ~11.5 million scholarly articles in engineering and computer science.

PQAI

Pros

Input Query

Google Patents requires the inventor to create complex keyword search strings, which can be time-consuming. Instead, PQAI understands the query in plain English, allowing for a faster and more seamless search .

Time Efficiency

PQAI is a real timesaver because:

  • Its AI fetches more relevant results.
  • Searchers can nudge the AI to bring more relevant results by pressing the “More Like This” button.
  • PQAI shows matching text from patent documents to make it easy to choose which document to read further.
  • In addition, the text matching is not just literal word-by-word, but its AI understands the intent of the search query. The AI also makes sure to include various synonyms of keywords present in the search query for a more inclusive search.

Security and Confidentiality

PQAI ensures your privacy by never logging any of your search queries or results.

Combinational Prior Art Search

PQAI allows the users to conduct a combinational prior art search. In such a search, PQAI shows a group of patents that indicate the inventive concept is not novel. Such a search helps the inventors to refine their ideas, reducing the probability of Section 103 patent rejections.

Cons

  • PQAI is an evolving collaborative initiative; it requires a community effort to create a world-class patent search portal to accelerate innovation and improve patent quality.

Lens

Lens is a patent search engine that provides free access to patents published across various technologies. Its search interface is exhaustive and complex—not an excellent choice for inventors.

Pros

  • Offers a search service for over 140 million patent records and other academic datasets.
  • Offers extremely fine-grained advanced search filters.
  • Offers a search in pending patent applications as well.

Cons

  • The interface is quite exhaustive with tons of search filters, making it more suitable for expert patent professionals who work on complex patent projects like landscaping.

PATENTSCOPE 

Offered by the World Intellectual Property Organization (WIPO), PATENTSCOPE is a worthy contender as a Google Patents replacement. Overall, PATENTSCOPE provides an impressive selection of resources that will help you stay up-to-date on the latest trends in technology innovation and licensing agreements.

WithPATENTSCOPE, you can narrow your patent searches using meta-level filters in over 100 million patents globally. It’s a search engine technology that’s available for free around the world. 

It’s a one-stop shop for patent search and research, offering a wide range of features, such as detailed patent descriptions, abstracts, images, and citations.

Pros

  • Allows searching for patents in languages other than English.
  • Provides instant translations of non-English patents.
  • Offers a chemical structure-based search. 
  • Allows downloading the results in Excel format.
  • Educates readers with tips and tricks to make the most of the search engine.

Cons

  • Requires the user to create a complex combination of fields to conduct the search.

USPTO

While widely used, there  are many weaknesses to the USPTO search interface, including a lack of filters for specific topics, limited results based on keyword relevance, and slow loading times. These problems can make it challenging to find the information you’re looking for quickly and efficiently.

Additionally, the data presented in the search results is comprehensive, but challenging to understand or use. The tables often contain too many columns and rows, which makes it difficult to focus on specific information.

 

Pros

  • This database contains more than 1 million patent records as of June 2018.
  • Allows searching in patents as well as patent applications using ~50 different fields.

Cons

  • Its search interface is complex and highly outdated. 
  • There’s no mobile app currently available. 

Espacenet 

Espacenet is a European patent database with over 130 million patents, including both granted and pending patents. 

Espacenet offers free access to information about inventions and technical developments from 1782 to today. It was developed by the European Patent Office (EPO) together with the member states of the European Patent Organization. Most member states have an Espacenet service in their national languages and access to the EPO’s worldwide database, most of which is in English.

Pros

  • Allows high-quality instant translation of non-English patents for over 30 languages.
  • Allows users to jump between results to review technical drawings for a quicker search.
  • Gives insightful data about inventors, applicants, and technology fields.

Cons

  • Requires users to create complex keyword search strings.

The Verdict

There are a few different alternatives to Google Patents that you may want to consider, but PQAI offers a very promising option. The PQAI search has many unique features to make it easier to use, compared to traditional search engines:

  • PQAI’s natural language interface enables inventors to describe their inventions using simple sentences.
  • PQAI displays only a limited set of results per search query. This makes the search process more straightforward and less time-consuming. Searchers may ask the system for “more like this” or “more like the save” to get more documents relevant to the search query.
  • PQAI guides researchers and helps you avoid Section 103 rejections.
  • Users can save relevant documents and incorporate them into a report for sharing with a patent attorney or evaluator to make the patent prosecution process more efficient.
  • PQAI protects users’ privacy by not storing searches after each session, instead saving documents and reports on the user’s computer.

Patent Quality through Artificial Intelligence (PQAI) is MORE THAN Just a Search Engine

PQAI is a not-for-profit organization that develops an open-source AI-based library of software components to speed innovation and boost patent quality

The flexibility and innovation of open-source software makes PQAI a leader in the patent ecosystem.

Committed to security, privacy and unbridled innovation, PQAI empowers all inventors with sophisticated IP tools to fuel their success and creativity.

Team PQAI has already built five applications to ease your patent search:

  1. Prior art search engine: Uses AI to pick limited most relevant prior art references. It also offers a combinational prior art search to reduce the probability of Section 103 patent rejections. Allows users to save results, generate prior art reports for patent prosecution discussion with a patent attorney, get insights on patenting trends and  more.
  2. CPCs look up: Lists the IPC/CPC codes based on the technology described in the search query.
  3. Art unit predictor: Tells the art unit the invention may fall into at the patent office. Different art units have different patent rejection statistics; applicants may wish to describe the inventive technology in a way that aligns it with a favorable art unit.
  4. Concept extractor: An invention might encompass multiple technical concepts. For instance, a coffee pot that keeps your coffee warm will have a temperature sensor, heating element, etc. While conducting a prior art search, you might want to get more ideas for  encompassing concepts.
  5. Keyword generator: Let’s say the main keyword is beer. Related keywords could be beverage, malt beverage, fermented beverage, alcoholic beverage, wine, drink, etc. While conducting the prior art search, this generator provides additional related keywords for a more comprehensive search.

The open-source nature of the PQAI project promises many more applications today and in the future to help patent professionals, corporations, and inventors. Visit PQAI to learn more.

3 Ways Inventors Can Leverage Prior Art to Fine-Tune Their Ideas

Ways Inventors Can Leverage Prior Art to Fine-Tune Their Ideas

You’ve come up with a brilliant invention and are ready to patent it. The first step in the process is to conduct a prior art search to ensure that there is no existing patent for your idea. The relation between your invention and the prior art will determine whether the Patent Office awards your patent or fights issuance with rejection after rejection.

But you might not have considered using the prior art in an iterative process to refine the invention. Whether you find no prior art, some prior art, or anticipatory prior art, you can use the prior art to improve the invention and increase your chances of obtaining a patent.

Here are three ways inventors can leverage prior art to fine-tune their ideas and improve the quality of their enterprise’s patents.

Creatively Interpreting and Using Prior Art Search Results

PQAI stands for Patent Quality Artificial Intelligence. This simple, AI-powered search engine provides your innovation’s ten extremely relevant prior art references.

After you run your innovative idea through the PQAI engine, for inventors and other non-attorneys, the search results fall into one of three categories:

  1. No relevant results
  2. Relevant results that anticipate your invention
  3. Relevant results that are close but not close enough to anticipate the invention

Regardless of which category your results fall into, the results give you valuable information about your next steps. Importantly, this is not a binary decision to either patent or abandon the idea. Instead, you can patent the invention or brainstorm and refine the idea to get onto the path to a patent.

In this article, you will learn a course of action for each category of results.

No Relevant Results Means Your Search May Need Refocusing

You face a dilemma if you do not get any relevant prior art references in a patent search. Your invention might genuinely be novel. If this is the case, your prior art search results might help the inventors develop new or related inventions to help your enterprise occupy the new field you have discovered.

Conversely, it could be that some relevant prior art does exist, but your search was flawed and failed to find those results. If you think the search results seem strange or inaccurate, you can refocus your search. In PQAI, you do this by:

  • Using the Concept Extractor application in PQAI to identify any missing or misinterpreted concepts by describing those missing concepts in greater detail
  • Identifying potentially relevant results and using the option in PQAI to show “more of these.”
  • Saving or opening potentially relevant results to nudge PQAI toward more results like those you viewed or saved.
  • Looking up CPCs with PQAI to ensure the system has searched for the correct concepts relating to the invention.

For example, suppose you invented a coffee machine with a sensor to turn off the burner when the coffee carafe is empty. Then, as you review your search results, you might find that they go toward coffee machines that turn off the heating element that boils the water for the brewer when the water tank is empty.

Source: PQAI Prior Art Search Engine

These results are not relevant to the subject invention. However, to find out if there are closer prior art references, you can use the tools built into PQAI to refocus the search.

Relevant Results With Existing Patents Can Spur Innovation

Suppose you run your search and find a prior art reference that precisely describes your invention. You might think these search results spell the end of your invention. But you can leverage these search results to continue innovating in the area and potentially create something new to be patented.

You can look at the later patents that cite your invention’s prior art reference and see how the concept has evolved. These forward citations might not necessarily tie directly to the original concept. They might have cited the references for some ancillary point. But these forward citations can describe one evolutionary path the invention took.

Within PQAI, you can use the save/open results functionality to nudge the system to find other related patents. This chain of patents might not have cited the original prior art reference, but they may illustrate similar concepts. Based on the references from forwarding citations and nudged results, you may find areas without prior art where you can tweak the invention.

Another option is to conduct a Google search to identify any products that ended up using the invention. This option can provide you with concrete applications for the invention that you might improve upon.

In the other direction, you may see that other enterprises have thoroughly covered the invention. So you could drop the idea altogether and move on to a different idea.

For example, suppose you invented a mobile phone with multiple cameras on one side of the phone. You could find results showing the exact concept and ideas built on it with 3D or infrared cameras.

Source: PQAI Prior Art Search Engine

Following the prior art chain that cited the closest reference or described concepts similar to the nearest reference, you can decide whether you have the space to tweak the idea or need to drop the idea.

Relevant Results With No Anticipating Patents Opens the Door to a Broad Patent

If your invention appears novel and you think you got good search results, you sit in a solid position to get a broad patent. Patent lawyers often compare patents to plots of land. You want your plot of land to sit directly adjacent to your neighbor’s plot, with no gaps in between.

When you read through the prior art found in your search, you may find gaps between your invention and the patented inventions. You can expand your invention to include other embodiments or applications that surround your invention. This approach gives you broader patent protection and closes the gaps for a possible design-around.

You can nudge the search results in PQAI to ensure no on-point results by using the save/open trick described previously. This step will give you peace of mind that you have found all the relevant references.

Once you feel comfortable about your search results, you can save your results in PQAI and download the search report for review by a patent attorney.

Going back to the coffee maker example, suppose you only find references that cover turning off the boiling element, not the warmer under the carafe. You realize your original idea of sensing the weight to turn off the warmer was too narrow. You can now expand your invention to cover other ways of sensing an empty carafe, such as optical sensors that “see” when the carafe is empty and temperature sensors that sense when the carafe starts to burn.

Using Prior Art Search Results Iteratively

You should view prior art searches as part of the innovation process. The search can tell you something about your invention no matter what prior art references you find.

As you review search results, you should remember to:

  • Refine searches when results seem off.
  • Read the references with a critical eye to find possible areas to tweak or expand your invention.
  • Resist the temptation to become frustrated if the prior art anticipates your invention.
  • Talk to a patent attorney about what you can and cannot patent.

Not Just a Patent Search Engine, An Initiative, A Movement, A Hope

PQAI is a not-for-profit initiative focused on creating an open-source AI-based library of software components to accelerate innovation and improve patent quality. We believe that establishing an open-source forum of IP tools will drive critical changes in the IP ecosystem like what Linux did in computing. The PQAI search engine is one example of how such components can be assembled to create new and useful tools. The empowering of all inventors with advanced IP tools will drive more diversity and inclusion, which will accelerate the pace of innovation. PQAI believes in transparency and user privacy.

To learn more about PQAI and explore opportunities to improve system feedback, contact Sam Zellner, project lead for PQAI.

The Evolution of Data Access Tools for Patents

The Patent Office has many restrictions on the information it can disclose. In fact, through the 1990s, prior art searches could only be conducted in the Patent Office’s library. And the early version of the Patent Office’s database only used patent classification codes and did not allow full-text searching.

Things started to change in the early 2000s. The U.S. began publishing pending applications in 2001, opening the door to a wealth of information for inventors, litigators, and patent prosecutors. Gone were the days of hiring agents near the Patent Office to conduct your patent searches and pull your file wrappers. Instead, anyone with some training and patience could use the Patent Office’s database to obtain patent and patent application data.

We have now reached the next stage in the evolution of patent data tools. Developers have identified the strengths and weaknesses in the Patent Office’s interface. They have created proprietary and open-source tools to obtain, clean, and visualize patent data.

Using the Power of Open-Source Tools to Transform the IP Landscape

Open-source tools have several advantages over proprietary approaches, including:

  • Free to use.
  • Open access to source code.
  • Available for developers to change or incorporate the source code into new tools.
  • Royalty-free distribution or redistribution.
  • New business models for corporates.

These attributes encourage developers to adopt a standard instead of creating separate approaches. It also speeds development by allowing developers to “stand on the shoulders” of their predecessors.

Obtaining Patent Data

Formerly, patent data was located in silos in the Patent Office. However, for the past 20 years, the Patent Office has made this information available. Some of the open-source tools designed to obtain patent data include:

PQAI

PQAI stands for Patent Quality through Artificial Intelligence. This library of patent-related tools provides a next-generation prior art search engine. This search engine evaluates the search results and returns the top ten prior art references. In addition, the search engine trains itself to determine which results to return based on historical patent examination records.

PQAI Search Engine

PQAI promises to transform the IP landscape for inventors/enterprises, patent attorneys, and even patent examiners by delivering higher-quality search results. Conducting a search and reviewing mountains of search results takes time. Since PQAI only provides the most relevant prior art references, it provides more accurate, faster, and cheaper patent searches.

PQAI was initiated by the Georgia Intellectual Property Alliance (GIPA) and AT&T. GreyB contributed the algorithm, and InspireIP manages the application. As an open-source application, developers continue to improve PQAI. To review PQAI or contribute, you can access the files on PQAI’s GitHub.

PatZilla

The initiators of PatZilla call it “a modular patent information research platform and data integration toolkit.” Its primary feature is a search engine that pulls prior art references from the European Patent Office’s database. It also pulls from DEPATISnet, CLAIMS Direct, and depa.tech. In addition, PatZilla provides pdf, image, bibliographic data, and full test acquisition from these services.

PatZilla’s contributions to the evolving IP landscape include:

  • A user interface that allows efficient screening of multiple references.
  • Web-based collaboration for information sharing.
  • Adaptable API for integration into third-party systems.

Andreas Motl authored PatZilla. But many developers have contributed to PatZilla since its initial release in 2014. To view the files for PatZilla or to contribute, go to PatZilla’s GitHub.

phpIP

phpIP manages and dockets patents and other IP rights. The software was designed for inventors, enterprises, and IP law firms.

The system’s initiators sought to develop a software package that was flexible and easy to use. Unfortunately, most alternative packages were complicated and provided more features than necessary. As a result, most users paid for features they did not need and could not use the features they wanted.

phpIP was built on open-source software. It is changing the IP landscape by providing intuitive docketing and patent management tool. Notably, users can adapt the system to their specific needs. As they do, they can contribute to the overall improvement of the system.

You can view the documentation and source code files at phpIP’s Github.

Cleaning Patent Data | Open Refine

Not every user who works with patent data will need to clean it. But occasionally, you will have a large file of patent or patent application data that does not have the correct format for your use.

In the past, users have relied on Excel or Open Office to clean data. But this often requires the user to manually fix each cell or have the programming knowledge to write a macro to fix the cells automatically.

Open Refine is a tool that automates patent data cleaning. It is an open-source tool that Metaweb Technologies, Inc, developed. It was acquired by Google and released for open use in October 2012.

Open Refine provides automated data cleaning functions that can be applied to large patent data files. Some of the features that apply to patent data cleaning include:

  • Reformatting dates.
  • Separating inventors into different cells.
  • Repairing corrupted or missing characters.

This tool can improve the speed and accuracy of the review, analysis, and storing of patent data. To contribute to Open Refine, visit the GitHub page.

Visualizing Patent Data

Visualizations can help identify trends or patterns in the massive amount of patent data that may relate to your project. For example, you might benefit from a visualization of when patent applications were filed or which countries they were filed in.

Until recently, you would need to comb through a spreadsheet to spot patterns in the data. Now,  there are tools to turn patent data into visualizations, including:

Gephi

Gephi is a network visualization platform that can create graphs showing relationships between patents or patent applications. It is an open-source application that is free to use. Association Gephi authored the software, but many developers have contributed to it. 

Gephi can convert CSV or Excel files into data visualizations. This means you can import a file from The Lens or a cleaned file from Open Refine (both discussed above). Gephi will then create a visualization of the data.

This will change the IP landscape by revealing obscure or hidden patterns in the data. For example, you can visualize the number of pending applications in the data file that belong to each assignee.

To view the source code for Gephi or participate in its development, visit the Gephi GitHub page.

Plotly

Plotly Chart Studio is an open-source platform that can be used to create interactive graphics. The open-source version of Plotly is cloud-based. This version is free to use. Plotly also offers enterprise versions for a fee.

Plotly creates graphs from data files generated through The Lens or Open Refine. Like Gephi, Plotly can help spot trends or patterns in the data. But unlike Gephi, the graphs in Plotly were designed to be interactive and shareable. This makes Plotly a valuable collaborative tool that will alter the IP landscape.

Plotly was developed by Plotly Technologies, Inc. You can help develop Plotly by reviewing the source code and documentation on Plotly’s GitHub.

Opportunities Ahead

The open-source nature of these tools almost guarantees that they will continue to develop and improve. To be a part of these opportunities, you can either use the software and provide feedback or you can collaborate with the developers to identify and create new features for these applications.

Get in touch with Sam Zellner, project lead for PQAI, to explore collaboration opportunities with PQAI.

6 Open-Source IP Tools That Inventors and Patent Professionals Must Know About

Open Source IP Tools for Inventors and Patent Professionals

The Evolution of Data Access Tools for Patents

The Patent Office has many restrictions on the information it can disclose. In fact, through the 1990s, prior art searches could only be conducted in the Patent Office’s library. And the early version of the Patent Office’s database only used patent classification codes and did not allow full-text searching.

Things started to change in the early 2000s. The U.S. began publishing pending applications in 2001, opening the door to a wealth of information for inventors, litigators, and patent prosecutors. Gone were the days of hiring agents near the Patent Office to conduct your patent searches and pull your file wrappers. Instead, anyone with some training and patience could use the Patent Office’s database to obtain patent and patent application data.

We have now reached the next stage in the evolution of patent data tools. Developers have identified the strengths and weaknesses in the Patent Office’s interface. They have created proprietary and open-source tools to obtain, clean, and visualize patent data.

Using the Power of Open-Source Tools to Transform the IP Landscape

Open-source tools have several advantages over proprietary approaches, including:

  • Free to use.
  • Open access to source code.
  • Available for developers to change or incorporate the source code into new tools.
  • Royalty-free distribution or redistribution.
  • New business models for corporates.

These attributes encourage developers to adopt a standard instead of creating separate approaches. It also speeds development by allowing developers to “stand on the shoulders” of their predecessors.

Obtaining Patent Data

Formerly, patent data was located in silos in the Patent Office. However, for the past 20 years, the Patent Office has made this information available. Some of the open-source tools designed to obtain patent data include:

PQAI

PQAI stands for Patent Quality through Artificial Intelligence. This library of patent-related tools provides a next-generation prior art search engine. This search engine evaluates the search results and returns the top ten prior art references. In addition, the search engine trains itself to determine which results to return based on historical patent examination records.

PQAI Search Engine

PQAI promises to transform the IP landscape for inventors/enterprises, patent attorneys, and even patent examiners by delivering higher-quality search results. Conducting a search and reviewing mountains of search results takes time. Since PQAI only provides the most relevant prior art references, it provides more accurate, faster, and cheaper patent searches.

PQAI was initiated by the Georgia Intellectual Property Alliance (GIPA) and AT&T. The algorithm was contributed by GreyB, and InspireIP manages the application. As an open-source application, developers continue to improve PQAI. To review PQAI or contribute, you can access the files on PQAI’s GitHub.

PatZilla

The initiators of PatZilla call it “a modular patent information research platform and data integration toolkit.” Its primary feature is a search engine that pulls prior art references from the European Patent Office’s database. It also pulls from DEPATISnet, CLAIMS Direct, and depa.tech. In addition, PatZilla provides pdf, image, bibliographic data, and full test acquisition from these services.

PatZilla’s contributions to the evolving IP landscape include:

  • A user interface that allows efficient screening of multiple references.
  • Web-based collaboration for information sharing.
  • Adaptable API for integration into third-party systems.

Andreas Motl authored PatZilla. But many developers have contributed to PatZilla since its initial release in 2014. To view the files for PatZilla or to contribute, go to PatZilla’s GitHub.

phpIP

phpIP manages and dockets patents and other IP rights. The software was designed for inventors, enterprises, and IP law firms.

The system’s initiators sought to develop a software package that was flexible and easy to use. Unfortunately, most alternative packages were complicated and provided more features than necessary. As a result, most users paid for features they did not need and could not use the features they wanted.

phpIP was built on open-source software. It is changing the IP landscape by providing intuitive docketing and patent management tool. Notably, users can adapt the system to their specific needs. As they do, they can contribute to the overall improvement of the system.

You can view the documentation and source code files at phpIP’s Github.

Cleaning Patent Data | Open Refine

Not every user who works with patent data will need to clean it. But occasionally, you will have a large file of patent or patent application data that does not have the correct format for your use.

In the past, users have relied on Excel or Open Office to clean data. But this often requires the user to manually fix each cell or have the programming knowledge to write a macro to fix the cells automatically.

Open Refine is a tool that automates patent data cleaning. It is an open-source tool that Metaweb Technologies, Inc, developed. It was acquired by Google and released for open use in October 2012.

Open Refine provides automated data cleaning functions that can be applied to large patent data files. Some of the features that apply to patent data cleaning include:

  • Reformatting dates.
  • Separating inventors into different cells.
  • Repairing corrupted or missing characters.

This tool can improve the speed and accuracy of the review, analysis, and storing of patent data. To contribute to Open Refine, visit the GitHub page.

Visualizing Patent Data

Visualizations can help identify trends or patterns in the massive amount of patent data that may relate to your project. For example, you might benefit from a visualization of when patent applications were filed or which countries they were filed in.

Until recently, you would need to comb through a spreadsheet to spot patterns in the data. Now,  there are tools to turn patent data into visualizations, including:

Gephi

Gephi is a network visualization platform that can create graphs showing relationships between patents or patent applications. It is an open-source application that is free to use. Association Gephi authored the software, but many developers have contributed to it. 

Gephi can convert CSV or Excel files into data visualizations. This means you can import a file from The Lens or a cleaned file from Open Refine (both discussed above). Gephi will then create a visualization of the data.

This will change the IP landscape by revealing obscure or hidden patterns in the data. For example, you can visualize the number of pending applications in the data file that belong to each assignee.

To view the source code for Gephi or participate in its development, visit the Gephi GitHub page.

Plotly

Plotly Chart Studio is an open-source platform that can be used to create interactive graphics. The open-source version of Plotly is cloud-based. This version is free to use. Plotly also offers enterprise versions for a fee.

Plotly creates graphs from data files generated through The Lens or Open Refine. Like Gephi, Plotly can help spot trends or patterns in the data. But unlike Gephi, the graphs in Plotly were designed to be interactive and shareable. This makes Plotly a valuable collaborative tool that will alter the IP landscape.

Plotly was developed by Plotly Technologies, Inc. You can help develop Plotly by reviewing the source code and documentation on Plotly’s GitHub.

Opportunities Ahead

The open-source nature of these tools almost guarantees that they will continue to develop and improve. To be a part of these opportunities, you can either use the software and provide feedback or you can collaborate with the developers to identify and create new features for these applications.

Get in touch with Sam Zellner, project lead for PQAI, to explore collaboration opportunities with PQAI.

Charting the Course Around Willful Infringement

willful

/ˈwɪlfʊl,ˈwɪlf(ə)l/

(of a bad or harmful act) intentional; deliberate.

“willful acts of damage”

There was a time courts treated willful infringement simply as an act of intentionally copying a patented invention. The onus was on the plaintiff to present convincing and clear evidence of an obvious infringement. Things changed; the alleged infringer now had to build a record to show that it did not act in bad faith. In 2016, the US Supreme Court issued an opinion wherein they stated that the Patent Act provides for enhancement of damages up to three times in case infringement is proved.

The courts have taken that even further in favour of the patent holder. Last month, the courts in Texas were of the opinion that willful infringement occurs when the infringement notice is not replied to. Holding Having silently watched the parade pass it by, Apple cannot now complain that the parade didn’t stop on its own to entertain them”, Apple has been ordered to pay damages to the tune of $308.5 million!

You can no longer ignore the infringement notice; it would be considered as a case of willful infringement!

Whether Willful Infringement or Not – Notice is here. Now What?

The logical first step would be to check the following:

  • Is the notice from a genuine patent holder or a patent troll? 
  • How strong is the subject patent? 
  • Does it really infringe? 
  • What’s the existing prior art? 
  • Has it genuinely been overlooked by the innovation team?
  • Was the innovation very easily available? 
  • Did it not occur to the development team that something so easily available might be patented?

Once you figure out if there is any merit in the notice, it’s time to decide which route to take – settle this dispute or fight the good fight in the court? 

A couple of things to remember 

  1. A settlement is almost always a good idea. Only if you see merit in the infringement notice you have received. 
  2. Taking the court route is extremely resource heavy. It will drain your enterprise of time, energy and of course, money. It might be worth depending on what’s at stake.

If you, after weighing all considerations, believe that an out of court settlement works better for you, there won’t be any need to respond to the notice. Make sure that your settlement agreement is air-tight when it comes to court claims against you though. 

If you do choose to go down the legal route through courts, you must reply to the notice.

Check Patent Strength

Wonder if there’s a quick and effective way to help you decide the route? A tool that helps you gauge the strength of the patent without even employing your IP team? The IP resources should be saved to look into the matter only if the subject patent appears real strong.

Imagine a search tool that can access any technical information out there – patent as well as non-patent literature. A search tool that has the most advanced algorithm to find you only the most relevant results. To top it all – conducting search on such a platform is a piece of cake. Sounds unbelievable?

We hear you, however it’s possible to create such a tool with the advances in AI capability to interpret the patent language. In fact we’ve taken the first step towards building such a tool. We call it PQAI – Patent Quality Through Artificial Intelligence.

PQAI is a collaborative initiative that aims to leverage AI to create a tool we just described. We envision – “A universal repository of humanity’s entire technical knowledge efficiently accessible to everyone and anyone.”.

At the moment, PQAI is far from perfect but it’s definitely promising. We say that with confidence because of the results that various validation test cases we have run have given us. A few months ago, we identified some IPR cases and ran the subject patents through PQAI. In some of these cases we observed that PQAI was able to find the prior art cited by the sued party to invalidate the subject patent. We’re sharing one of those cases with you right here:

Facebook VS Villmer LLC Inc

Facebook had received an infringement notice over Oculus Go from Villmer. We used PQAI to run a prior art search to see if it could have helped locate prior art. It gave us results that matched the prior art that was submitted to have the patent invalidated. 

The patent that was the subject matter of the litigation bore the number US9618747. The patent broadly pertains to head-mounted displays. Villmer’s infringement allegations focused on Facebook’s head-mounted virtual reality (VR) devices, the Oculus Go and the Oculus Quest. The ’747 Patent was valid, enforceable, and was duly issued in full compliance with

US law. 




A picture of Oculus Go by Facebook.

 A drawing from Villmer’s patent

We picked up claim 1 from patent number US9618747 and pasted it in the PQAI query box. Here’s a screenshot of one of the prior art that Facebook listed during IPR to have ‘747 invalidated.

Prior art snapshot from IPR document

Here’s the prior art PQAI spotted against our search for Claim 1 from ‘747. The spotted patent bears no. US2012005053A1, one of prior arts listed by Facebook: 

Running a query through PQAI

PQAI is a search tool that is a work in progress. We would like to see it get to a stage where prior art like this ranks higher and results are even more efficient. And we are constantly working toward perfecting the system. 

Get Involved – Join the PQAI Initiative

PQAI is a collaborative initiative  to build an AI-powered prior art search tool accessible to all involved. We would love all stakeholders to come forward and join the initiative. You as a corporate can contribute too! The question is why would you….

Well with PQAI, we aim to create a more stable and transparent patent ecosystem for corporations. We believe it will make R&D investments more predictable, increase confidence in patents, and make risks and opportunities more visible. We could use all the resources we can get in making this initiative a success. You could be part of this revolution by funding the development of the tool. In return, you will give to yourself, and the world at large a system to file better patents. 

Disclaimer – “The statements and views expressed in this post are intended for general informational purposes only, and do not constitute legal advice or a legal opinion.”

The Top 20 AI Inventors and their Most Cited Patents

If our era is the next Industrial Revolution, as many claim, AI is surely one of its driving forces.

 – Fei-Fei Li

AI is no more limited to sci-fi movies. The endeavor to replicate or simulate human intelligence in machines has led to AI being mainstream in the last decade. AI has left a lasting impact on all our lives.  From being a figment of our imaginations to becoming an intrinsic part of our every day, the AI revolution is real and is here to stay. 

We looked back at 2020 and put together a list of the top 20 inventors to Artificial Intelligence. 

Top 20 AI inventors
  1. Sarbajit Rakshit

An Application Architect and seeker of solutions, Sarbjit Rakshit is an IBM Master Inventor with a degree in mechanical engineering from the Indian Institute of Engineering, Science and Technology.

Source – Forbes

He was awarded 163 U.S. patents in 2019, the highest ever awarded to a citizen of India in a single year. His patent portfolio contains 359 patents in Artificial Intelligence globally belonging to 271 unique patent families.

The most valuable patent in Sarbajit’s portfolio is US20160070439A1Electronic commerce using augmented reality glasses and a smartwatch. This patent family is the most cited (47 times), by companies Ariadne’s Thread (USA) Inc., Microsoft Technology Licensing Llc, Siemens Ag, Ebay Inc, Lucyd Ltd.

Source – US20160070439A1

Source – Verdict

Before we look at the rest of the list, here’s an interesting insight. 11 of the top 20 AI inventors are either currently at StradVision or have worked there previously. 10 of these inventors are co-inventors on a patent. Not just any patent, it’s their most cited patent. Let’s find out what StradVision does and what their most cited patent is about.

StradVision is a fairly new company, founded in late 2014. Their goal is to bring powerful and safe ADAS (Advanced driver-assistance systems) & self-driving technology to the masses. StradVision’s technology utilizes a novel perception algorithm allowing autonomous vehicles to reach the required level of safety, accuracy, and driver convenience. This is achieved through safe & reliable real-world object detection, tracking, segmentation, and classification. They have an auto labeling system that produces training data with minimal human input, and a semi-supervised learning-based training tool, enabling autonomous vehicles to detect and perceive environments in real-time.

StradVision - AI Assisted Driving For everyone

Source – StradVision

The most cited patent for these 10 inventors is US10169679B1. The patent is for – “Learning method and learning device for adjusting parameters of CNN by using loss augmentation and testing method and testing device.

StradVision

Source: StradVision

Yongjoong Kim, Woonhyun Nam, Sukhoon Boo, Myungchul Sung, Donghun Yeo, Wooju RYU, Taewoong Jang, Kyungjoong Jeong, Hongmo Je, Hojin Cho are co-inventors on the  said patent.

The said patent family has been cited 27 times, by companies Didi Res America Llc, Stradvision Inc, and Beijing Didi Infinity Technology. The patent’s geographical coverage extends to the United States, China, Japan, and Korea. 

  1. Wooju Ryu

Wooju Ryu is a Korean inventor and holds a master’s degree in Computer Engineering from Pohang University of Science and Technology. 

Wooju Ryu AI Inventor

Source: Twitter

He is presently an Algorithm Engineer at StradVision and works in areas of Deep Learning, Computer Vision, ADAS, Text Recognition and Automatic Driving. He has been associated with Intel, Olaworks, and Samsung as a Senior Researcher between 2007 and 2016. 

Wooju Ryu - Technology wise Patents Distribution

His patent portfolio consists of 831 patents in the AI domain globally, which belong to 267 unique patent families.

  1. Woonhyun Nam

Woonhyun Nam is a Korean inventor and holds a bachelor’s degree in Computer Science Engineering and a Doctor of Philosophy (Ph.D.) Computer Science and Engineering from the Pohang University of Science and Technology.

Woonhyun Nam - AI inventor

Source: ResearchGate

He is presently the Director, Lead of Algorithm Engineering at StradVision, Inc. His work profile is deeply seated in AI, with him being responsible for engineering, researching, investigating, and deploying algorithms across company products and services. 

Woonhyun Nam - Technology Area Patent famility count

His portfolio consists of 826 patents in the AI domain globally which belong to 266 unique patent families. Most of his inventions are in the field of Instruments Technology.

  1. Hongmo Je

Hongmo Je is a Korean inventor and holds a degree in Computer Science from the Pohang University of Science and Technology. 

Hongmo Je AI Inventor

Source: Crunchbase

Presently, he is the CTO of Stradvision and leads the RnD Integration/Engineering Team developing camera-based perception SW stack for ADAS/Autonomous Driving applications. He has previously been the Engineering Manager at Intel and the head of RnD at Olaworks. 

HongMo Je - Technology wise Patent family count

Hongmo Je’s patent portfolio consists of 824 patents in the Artificial Intelligence (AI) domain globally which belong to 264 unique patent families. He holds 256 patents in the Instruments domain. 

  1. Donghun Yeo

Donghun Yeo is a Korean inventor and holds a bachelor’s degree in Computer Science and a Ph.D. in Computer vision from Pohang University of Science and Technology. 

Donghun Yeo - AI Inventor

Source: NIST

Yeo is presently a Senior Researcher at the Hana Institute of Technology. Previously, he was an algorithm engineer at StradVision.

Yeo’s patent portfolio consists of 824 patents in the Artificial Intelligence domain globally belonging to 264 unique patent families. The major chunk of his portfolio consists of innovations in Instrument Technology (255). 

  1. Myungchul Sung

Myungchul Sung is a Korean inventor and holds a master’s degree in Computer Science Engineering from the Pohang University of Science and Technology. He is an Algorithm Engineer at StradVision. 

He holds 824 patents in the Artificial Intelligence domain globally which belong to 264 unique patent families. The largest chunk of his patent portfolio is innovations in the Instruments Technology domain, amounting to 255. 

  1. Yong-Joong Kim

Yong-Joong Kim is a Korean inventor with a master’s degree in Computer Science from Yonsei University. He is presently an algorithm engineer at Stradvision. In the past, he has been a researcher at Pohang University of Science and Technology, and an IT coordinator at the National Institute for International Education. He has interned at the MARG Lab at Seoul National University.

  1. Taewoong Jang

Taewoong Jang is a Korean inventor with a bachelor’s degree in Physics & Math, who graduated Magna Cum Laude from the Pohang University of Science and Technology. He was an Algorithm Engineer at StradVision and is now a Software Engineer at Coinone. 

He holds 824 patents in the Artificial Intelligence domain globally across 264 unique patent families. The majority of his patent portfolio (255 patents) are innovations related to Instruments Technology. 

  1. Kyungjoong Jeong

Kyungjoong Jeong is a Korean inventor who is an Algorithm Engineer at Stradvision. He graduated from the Ulsan University as an Electrical Engineer as the Dean’s Honoured Graduate. He has previously been at Samsung Techwin and a Researcher at POSTECH from where he earned his Master’s degree. His research interests are in Deep Learning, Computer Vision, Machine Learning.

Kyungjoong Jeong’s patent portfolio has 824 patents in the Artificial Intelligence (AI) domain globally which belong to 264 unique patent families. 255 of these patents are innovations in the field of Instruments Technology. 

  1. Hojin Cho

Hojin Cho is a Korean inventor and holds a degree in Computer Science Engineering and Doctor of Philosophy (Ph.D.) Image Processing, Computer Graphics, and Computer Vision from the Pohang University of Science and Technology. He is an Algorithm Engineer at StradVision.


His portfolio consists of 824 patents in AI  belonging to 264 unique patent families, of which 255 are in the sub-domain of Instruments Technology.

  1. Sukhoon Boo

Sukhoon Boo is a Korean inventor associated with StradVision Inc. His portfolio consists of 824 patents in AI  belonging to 264 unique patent families, of which 255 are in the sub-domain of Instruments Technology.

  1. Hak-Kyoung Kim

Hak-Kyoung Kim is a Korean inventor and is an algorithm engineer affiliated with Stradvision Inc. 

His portfolio consists of 758 patents in Artificial Intelligence globally, belonging to 251 unique patent families. He has 242 innovations in the domain of Instruments Technology.

The most valuable patent in Hak-Kyoung Kim’s portfolio is US10229346B1 – 

Learning method, learning device for detecting object using edge image and testing method. This is his most cited patent having been cited 13 times. The patent’s geographical coverage is in the United States, China, Korea, and Japan.

SourceUS10229346B1

  1. Kye-Hyeon Kim

Kye-Hyeon Kim is a Korean inventor and holds a bachelor’s degree in Computer Science and a Ph.D. in Computer Science (Machine Learning) from the Pohang University of Science and Technology.

Currently, he is the Chief Research Officer at Superb AI Inc. He has previously been associated with StradVision as an Algorithm Engineer, SK Telecom as a Research Scientist, Intel, and Samsung as a Senior Software Engineer. 

He holds 754 patents in the Artificial Intelligence domain globally which belong to 251 unique patent families. The largest chunk of his innovations is in the domain of Instruments Technology (242).

The most valuable patent in his portfolio is US10229346B1, same as Hak-Kyoung Kim. They are co-inventors with a few more inventors on this patent.

  1. John M Ganci Jr

John M Ganci Jr is an American inventor affiliated with IBM.  His patent portfolio has 223 patents filed globally which belong to 145 unique patent families. He holds 102 patents in the Instruments Technology domain.

John M Ganci Jr.’s most cited patent is US20160070439A1, same as Sarbajit Rakshit. They are co-inventors on this patent with a few others.

  1. Craig Trim

Craig Trim is an American inventor and holds a Bachelor’s degree in Computer and Information Sciences from Cal Poly Pomona and a Master of Science, MS, Data Analytics from Capella University.

Source – TheOrg

He is currently with Causality Link as a Senior Engineer. His past experiences include being at IBM as a Lead Data Scientist and Dristi as a CTO.

Trim’s patent portfolio consists of 223 patents in the AI domain globally which belongs to 144 unique patent families. He holds 116 patents in the Instruments Technology domain. 

Craig’s most cited patent is US20160070439A1, Craig is a co-inventor on this with Sarbajit Rakshit, John Gangci and few others.

  1. Corville O Allen

Corville Allen is an American inventor and holds a degree in Computer Science, Mathematics from Lona College. He has 17 years of experience in Enterprise Software Development including web-based software, Application Server infrastructure, Business Application Integration, and Cognitive Systems. He is a Senior Technical Staff Member and Master Inventor, 5-time North Carolina Inventor of the Year at IBM. 

Source: IBM News Room

His specialities include Application Integration, API Development, Agile Methodologies, SDLC, WebSphere, Connectivity, Architecture.

His patent portfolio consists of 232 patents globally which belong to 142 unique patent families. He holds 120 patents in the domain of Instruments Technology.

Allen’s most valuable patent is US9369488B2Policy enforcement using natural language processing. The said patent family has been cited 119 times by company Onetrust Llc. The patent’s geographical coverage extends to the United States and China.

The core idea of the patent is to automatically identify if the user is violating the “terms of use” policy for devices like computers. For example: one example scenario, the user may attempt to use the device camera to take a photograph of an object within a physical location governed by the term of use policy document. Based on the procedure disclosed in teh patent  the user’s computing device may then take an appropriate action, e.g., policy enforcement, restricting or disabling functionality, alerting or warning the user to non-compliance, or the like.

  1. Martin G Keen

Martin Keen is an American inventor and with a degree in Computer Science from Southampton Solent University. He has been associated with IBM as a Technical Content Creation Leader & Video Production Leader.

Source: The Marketplace Podcast

Martin is an IBM Master Inventor and was conferred the Honorary award in 2016 by IBM. He holds over 200 patent applications issued specializing in areas such as big data, cognitive systems, mobile devices, and predictive analytics. Martin is a Technical Content Creator Leader including the development of dozens of published books. He is also a Videographer and Video Production Lead specializing in corporate video creation and online learning course development. 

His patent portfolio has 201 patents filed globally which belong to 138 unique patent families. He holds 90 patents in the domain of Instruments Technology. 

The most valuable patent in Martin Keen’s portfolio is US9473819B1 – Event pop-ups for video selection. The said patent family has been cited 16 times by companies IBM, Sony Interactive Entertainment Llc, Amazon Tech Inc, Dish Network Llc.

Source – US9473819B1

  1. Jeremy Fox

Jeremy Fox is an American inventor who holds a degree in BBA, Computer Information System from the University of Texas at El Paso. He has been associated with IBM since 2001. He has been accorded the title of Master Inventor at IBM. 

Jeremy has also been serving as the IBM Commerce IDT Chair for over 3 years.

His patent portfolio consists of 128 patents in AI globally belonging to 110 unique patent families. 68 patents have been filed in the domain of Instruments Technology. 

The most valuable patent in Jeremy Fox’s portfolio is US9826500B1 Preventing driver distraction from incoming notifications – cited 8 times by Nocell Technologies Llc.

Source – MyPolice #LeaveThePhoneAlone

Don’t we agree –  those smartphone notifications while driving can be dangerously distractive? Jeremy Fox’s thought process behind this patent is quite appreciable. His ingenious idea is to adjust the intensity of notification alerts based on the driving conditions is remarkable. For example: changing loud beep to just a vibration alert for a certain type of notification. A few examples of conditions include driving:

  • in fair/poor/good weather
  • during day/night
  • familiar/unfamiliar route
  1. Yasuaki Yamagishi

Yasuaki Yamagishi is a Japanese inventor who is currently a Senior Research Scientist at Sony Corporation.

His patent portfolio there consists of 614 patents globally which belong to 104 unique patent families. He has 99 patents in the domain of Electronics Communication Technique.

His patent US10178148B2Content supply device, content supply method, program, and content supply system – is his most cited (13 times), by Sony Corporation, Saturn Licensing LLC. The patent’s geographical coverage extends to the United States, Brazil, India, China, and Russian Federation.

  1. Joydeep Ray

Joydeep Ray is an American inventor with a master’s degree in Computer Engineering from the Carnegie Mellon University. He is a Graphics Architect at Intel Corporation and has previously been associated with AMD as an MTS Design Engineer, Standard Performance Evaluation Corporation as a Technical Representative in CPU Sub-committee, Carnegie Mellon University as a Research Assistant, and IBM as a Design Engineer.

Ray’s patent portfolio has 293 patents in the Artificial Intelligence domain globally belonging to 84 unique patent families. 77 inventions are related to instruments belonging to the Instruments Technology domain. 

His patent US10108850B1Recognition, reidentification, and security enhancements using autonomous machines – is his most valuable. It has been cited 10 times and has geographical coverage in the United States and China. 

Let’s Sum it Up

It was interesting to note that most inventors among the top 20 AI inventors across the globe are Korean. 12 out of 20 are either working at StradVision or have worked at StradVision in past. It’s intriguing to know what StradVision is upto. There is a commonaliity in many of these inventors’ most cited patents as well. It’s the object recognition in a video.

InventorCountry of OriginPresent Place of WorkPast Places of Work
Sarbajit RakshitIndianIBMN/A
Wooju RyuKoreanStradVision Inc.Intel, Olaworks, Samsung
Woonhyun NamKoreanStradVision Inc. N/A
Hongmo JeKoreanStradVision Inc.Intel, Olaworks
Donghun YeoKoreanHana Institute of TechnologyStradVision Inc.
Myungchul SungKoreanStradVision Inc.N/A
Yong-Joong KimKoreanStradVision Inc. N/A
Taewoong JangKoreanCoinoneStradVision Inc.
Kyungjoong JeongKoreanStradVision Inc. Samsung Techwin
Hojin ChoKoreanStradVision Inc. N/A
Sukhoon BooKoreanStradVision Inc. N/A
Hak-Kyoung KimKoreanStradvision Inc.N/A
Kye-Hyeon KimKoreanSuperb AI Inc. StradVision, SK Telecom, Intel and Samsung
John M Ganci JrAmericanIBMN/A
Craig TrimAmericanCausality LinkIBM, dristi
Corville O AllenAmericanIBMN/A
Martin G KeenAmericanIBMN/A
Jeremy FoxAmericanIBMN/A
Yasuki YamagishiJapaneseSony CorporationN/A
Joydeep RayAmericanIntel CorporationAdvanced Micro Devices Inc., IBM

What kindled your interest in this article. Are you currently working on any AI projects?

Since you showed an interest in this article, we wish to share an AI-based initiative with you. It’s called Patent Quality through Artificial Intelligence. The initiative is focussed on inventors and the core value that drives the initiative is “Prior Art Search for Everyone”. At PQAI, we studied patent rejection stats. We observed that most patents receive 102/103 type rejections. This means the invention described in the patent is either not new or obvious based on a combination of one or more previous inventions/literature. Many inventors apply for patents without conducting a thorough prior art search. Usually, this is because there is a lack of budget or patent searching skills. Also, it’s quite difficult to search for non-patent literature while performing a prior art search. These reasons triggered in us an urge to develop an inventor friendly prior art search engine. And what better than AI to turn to for help?

If you feel the pain inventors go through on receiving a patent rejection, we urge you to join the initiative and contribute the best way only you can!

Women’s History Month: Honoring 15 Ingenious Female Inventors

Necessity is the mother of invention.” – Plato

On May 5, 1809, Ms. Mary Dixon Kies became the first woman to receive a patent in the United States of America for her technique of weaving straw with silk. Women, from time immemorial, have been innovating and making breakthroughs in the technological world. From windshield wiper to coffee filter paper, women have contributed significantly with their inventions to make this world a better place.

March is observed as the women’s history month, to reflect on the often-overlooked contributions of women in history. To mark women’s history month, we honor the contributions of these 15 female inventors who have been driving innovation.

#1. Catia Bastioli

Ms. Catia Bastioli is an Italian inventor, chemist, researcher, and entrepreneur. She holds a degree in pure chemistry from the University of Perugia, Italy. Ms. Bastioli also attended the School of Business Administration (“Alti Potenziali Montedison”) at the Bocconi University in Milan. Ms. Bastioli is the CEO of Novamont S.p.A. She is also the president of Terna Spa of the Kyoto Club Association and of the Italian Technological Cluster of Green Chemistry SPRING and member of the Board of Directors of Fondazione Cariplo. 

Awards & Honors:

  • European Inventor of the Year Award in 2007 in the category “SMEs/research”
  • Honoris Causa Degree in Industrial Chemistry (2008, University of Genoa)
  • Honorary title of Knighthood (“Cavaliere dell’Ordine al Merito della Repubblica Italiana”), 2013
  • Honoris Causa Degree in Materials Engineering (2016, University of Palermo) 
  • Appointed as “Cavaliere del Lavoro” by the President of the Italian Republic in 2017
  • Honoris Causa Degree in Business Economics (2018, University of Foggia)
  • Honorary Doctoral Degree in Civil, Chemical, Environmental, and Materials Engineering (2019, University of Bologna)

Her most valuable patent is US5412005A for biodegradable polymeric compositions based on starch and thermoplastic polymers. 

15 Ingenious Female Inventors - Catia Bastioli

Source – EPO

Ms. Bastioli’s patent portfolio has 1291 patents globally which belong to 186 unique patent families. She is an individual inventor of 4 and a co-inventor in the rest of the 182 core patents. 

#2. Esther Sans Takeuchi

Dr. Esther Sans Takeuchi is an American inventor. She completed her graduation from the University of Pennsylvania. She completed her Ph.D. in organic chemistry from Ohio State University, under the direction of Dr. Harold Shechter in 1981. She has been a Professor at the State University of New York at Buffalo since September 2007. Prior to this, she has worked with Electrochem as a Chief Scientist and at Greatbatch Inc. for more than two decades as a Director of Research & Development. 

She has also served as a postdoctoral research associate in electrochemistry, first at the University of North Carolina at Chapel Hill from 1982 to 1983, and then at the State University of New York at Buffalo from 1983 to 1984.

Dr. Takeuchi is a member of the US National Academy of Engineering. After 40 years in industry and academia, she continues to work at the forefront of battery technology innovation

Awards & Honors:

  • National Medal of Technology and Innovation in 2010 awarded by President Obama
  • European Inventor Award in 2018

The most valuable patent in Ms. Takeuchi’s portfolio is US4964877A for a non-aqueous lithium battery. 

#WomenHistroryMonth 

Dr.Esther Sans Takeuchi

Source – EPO

Ms. Takeuchi has 584 patents globally which belong to 153 unique patent families in her patent portfolio. She is an individual inventor of 5 and a co-inventor in the rest of the 148 core patents.

#3. Joy Mangano

Ms. Joy Mangano is an American inventor and entrepreneur. She was the president at Ingenious Designs LLC. She completed her graduation in business administration from Pace University.

In 1990 after growing frustrated with ordinary mops, Ms. Mangano developed her first invention, the Miracle Mop. It is a self-wringing plastic mop with a head made from a continuous loop of 300 feet (90 meters) of cotton that can be easily wrung out without getting the user’s hands wet. David O. Russell directed an Oscar-nominated movie based on her life, Joy. Ms. Mangano has also written a best-selling book, Inventing Joy which she says is for those who want to build a brave and creative life.

Female Inventors - Joy Mangano - Self Wringing Mop

Source – The Inc

Awards & Honors:

  • Named the Long Island Entrepreneur of the Year by Ernst & Young in 1997
  • Ranked number 77 on Fast Company’s list of the 100 Most Creative People in Business in 2009
  • Included in Fast Company’s list of the 10 Most Creative Women in Business in 2010

The most valuable patent in Ms. Mangano’s portfolio is US5722260A for reversible jewelry clasp for necklaces and/or bracelets. 

Source – US Patent – 5722260A
Source – PurePearls

Ms. Mangano’s patent portfolio consists of 121 patents globally which belong to 68 unique patent families. She is an individual inventor of 55 and a co-inventor in the rest of the 13 core patents.

 

#4. Helen Lee

Dr. Helen Lee is a medical researcher. She obtained her Ph.D. in biology, microbiology, and parasitology from Cornell University. Dr. Lee is the Associate Professor in Medical Biotechnology at the University of Cambridge. She is also the President and CEO of Diagnostics for the Real World Ltd (DRW), Sunnyvale, USA, and its wholly-owned subsidiary, DRW-Europe, Cambridge, UK. She has also worked with Abbott Laboratories from 1991 to 1995, as a General Manager, Probe Diagnostics Business Unit.

Awards & Honors:

  • National Honor Society of Sigma Xi, 1967
  • Who’s Who of American Men in Science, 1970
  • The Entrepreneurial Award (Abbott Laboratories), 1988
  • The Phoenix Award (Abbott Laboratories), 1991
  • Finalist, YWCA Women of Achievement Award, 1994
  • Best Diagnostic Innovation Award (Medical Futures Innovation Competition), 2003
  • Lord Lloyd of Kilgerran Award (British Foundation for Science & Technology), 2005 
  • British Female Inventor in Industry Award, 2006
  • European Women of Achievement Award 2006
  • Asian Women of Achievement Award, 2007
  • Tech Museum of Innovation Award, 2007 
  • European Inventor Award, 2016
  • Appointed as a judge for the European Inventor Award, 2019
  • Recognized on the Times’ Science Power List in May 2020 

Her invention, the diagnostic kit SAMBA II is being repurposed for use in COVID-19 testing. The most valuable patent in Ms. Lee’s portfolio is US6521747B2 for haplotypes of the AGTR1 gene. 

SAMBA by Helen Lee

Source – EPO

Dr. Lee has 168 patents globally which belong to 37 unique patent families in her patent portfolio. She is an individual inventor of 4 and a co-inventor in the rest of the 33 core patents.

#5. Ann Lambrechts

Ms. Ann Lambrechts is a Belgian inventor. She is working with N.V. Bekaert S.A. as a global sales and product manager. Prior to this, she was working as the head R&D of Building Products at Bekaert.

Awards & Honors:

  • Winner of the European Inventor Award, 2011 in the category “Industry”
Female Inventors - Ann Lambrechts

Source – EPO

Her invention of mixing steel wire elements into concrete has improved the stability of structures where it is used and reduced building costs. This invention, which is patent no. US6235108B1 is the most valuable patent in her portfolio. Her invention increases the bending tensile strength of concrete by 32%, enabling more pioneering projects to be built.

Ms. Lambrechts patent portfolio has 232 patents globally, which belong to 28 unique patent families. She is an individual inventor of 10 and a co-inventor in the remaining 18 core patents.

#6. Ursula Keller

Dr. Ursula Keller is a Swiss inventor. She obtained her Ph.D. in engineering physics/applied physics from Stanford University.

Dr. Keller joined ETH Zurich as a professor of physics in 1993, where she leads the Ultrafast Laser Physics group. She currently serves as a director of the NCCR MUST (Molecular Ultrafast Science and Technology), an interdisciplinary research program supported by the Swiss National Science Foundation, bringing together 15 Swiss research groups in ultrafast physics and chemistry. She has published more than 330 peer-reviewed journal papers and 11 book chapters.

Awards & Honors:

  • Weizmann Women and Science Award, 2017
  • European Inventor Award, 2018 for laser technology in the category “lifetime achievement”
  • IEEE  (Institute of Electrical and Electronics Engineers) Photonics Award, 2018
  • IEEE (Institute of Electrical and Electronics Engineers) Edison Medal, 2019 
  • SPIE (the international society for optics and photonics) Gold Medal, 2020

The most valuable patent in Dr. Keller’s portfolio is US6834064B1 for the semiconductor saturable-absorber mirror technology used in mode-locking ultrafast solid-state laser systems. 

Source – EPO

She has 80 patents globally which belong to 22 unique patent families in her patent portfolio. She is an individual inventor of 2 and a co-inventor in the rest 20 core patents.

 

#7. Christine Van Broeckhoven

Dr. Christine Van Broeckhoven is a Belgian inventor. She completed her Ph.D. in molecular biology from the University of Antwerp. Ms. Broeckhoven is a Professor of Molecular Biology and Genetics at the University of Antwerp since 1995. 

Since 1983 she has had her own laboratory for molecular genetics at the University of Antwerp, and since 2005 is focussing her research on neurodegenerative brain diseases. She is an associate editor of the scientific journal Genes, Brain, and Behavior.

Dr. Broeckhoven has over 35 years of experience in molecular genetics research of neurodegenerative brain diseases such as Alzheimer’s disease, frontotemporal lobar degeneration, amyothrophic lateral sclerosis, lewy bodies disorders, and Parkinson’s disease.

Awards & Honors:

  • Belgian Quinquennial Prize of the Belgian National Science Foundation 
  • Potamkin Prize (The Potamkin Prize for Research in Pick’s, Alzheimer’s, and Related Diseases), 
  • The Arkprijs van het Vrije Woord 
  • European Inventor Award 2011.

The most valuable patent in Dr. Broeckhoven’s portfolio is EP561087B1 for a mutated form of the beta-amyloid precursor protein gene. 

Source – magazine.live

Her patent portfolio has 77 patents globally which belong to 20 unique patent families.

#8. Margarita Salas

Late Dr. Margarita Salas (30 November 1938 – 7 November 2019) was a Spanish inventor. Margarita had graduated from the Complutense University of Madrid with a B.A. in chemistry and obtained a Ph.D. in 1963. She started her career in the US-based laboratory of Nobel-prize winner Severo Ochoa. She returned to her native Spain in 1967 to establish the country’s first research group in the field of molecular genetics.

Dr. Salas led the breakthroughs that have since made DNA testing fast, reliable, and used in a wide range of applications.

Awards & Honors:

  • Carlos J. Finlay Prize, UNESCO, 1991
  • Medal of Principality of Asturias, 1997
  • National Research Award Santiago Ramon y Cajal, 1999
  • L’Oreal-UNESCO Award for Women in Science, 2000
  • Selected among the 100 women of the twentieth century that paved the way for equality in the XXI Century by the Council of Women of the Community of Madrid, 2001
  • Isabel Ferrer Award of the Generalitat Valenciana, 2002
  • Gold Medal of the Community of Madrid, 2002
  • Grand Cross of the Civil Order of Alfonso X, the Wise, 2003
  • International Prize for Science and Research Cristóbal Gabarrón Foundation, 2004
  • Gold Medal for Merit in Work, 2005
  • Medal of Honor of the Complutense University of Madrid, 2005
  • Award of Excellence granted by FEDEPE (Spanish Federation of Women Directors, Executives, Professionals, and Entrepreneurs), 2006
  • First Spanish woman to become a member of the National Academy of Science (United States), 2007
  • Gold Medal of the College of Veterinarians of the Principality of Asturias, 2009
  • Title of Honorary Ambassador of the Spain Brand, category of Science and Innovation, which fails Leading Brands of Spanish Forum with the approval of the Ministry of Foreign Affairs and Cooperation, 2009
  • Women Leader Award, awarded by the Rafael del Pino, Aliter and Merck Foundation, 2009
  • Award “An entire professional life” of the Mapfre Foundation, 2009
  • Chemistry Excellence Award, awarded by the General Council of Associations of Chemists of Spain, 2014
  • Medalla Echegaray, the highest award from the Spanish Royal Academy of Sciences, 2016
  • ManchaArte Award 2018, 2018
  • European Inventor Award Lifetime Achievement Award and Audience Award by European Patent Office, 2019

The most valuable patent in Dr. Salas’s portfolio is US5198543A for an improved method for determining the nucleotide base sequence of a DNA molecule. 

Source – The Conversation

Her patent portfolio has 80 patents globally which belong to 19 unique patent families.

#9. Marissa Mayer

Ms. Marissa Mayer is an American inventor. Marissa studied symbolic systems and computer science with an emphasis on artificial intelligence, receiving a B.S. degree in 1997 and an M.S. degree in 1999 at Stanford University. She is the co-founder of Sunshine Contacts. She has worked with some of the major corporate giants like Walmart, Yahoo, and Google. Ms. Mayer designed the search interface of Google’s home page. During her tenure at Google, Ms. Mayer helped create a number of patented inventions related to web-browsing software, including a program that searches saved articles.

Source – Business Insider

Ms. Mayer actively invests in technology companies, including crowd-sourced design retailer Minted, live video platform Airtime.com, wireless power startup uBeam, online DIY community/e-commerce company Brit + Co., mobile payments processor Square, home décor site One Kings Lane, genetic testing company Natera, and nootropics and biohacking company Nootrobox.

The most valuable patent in Ms. Mayer’s portfolio is US7096214B1 for a system and method for supporting editorial opinion in the ranking of search results. 

Source – US7096214B1

Ms. Mayer has 63 patents globally which belong to 14 unique patent families in her patent portfolio.

#10. Annegret Matthai

Ms. Annegret Matthai is a German inventor, working with Audi AG, Germany. She is involved and working on inventions related to the motor industry. The most valuable patent in her portfolio is DE102008004049A1 for a laminated glass unit for use as a windshield in a motor vehicle.

Source – DE102008004049A1

Ms. Matthai has 32 patents globally in her patent portfolio, which belong to 13 unique patent families.

#11. Ann Tsukamoto

Dr. Ann Tsukamoto is an American inventor with a Ph.D. in microbiology and immunology from the University of California. She is a stem cell researcher, who started her career with SyStemix in 1989.

With her husband, Professor Irv Weissman, as co-patentee, Dr. Tsukamoto’s patent for stem cell isolation was awarded in 1991. Their discovery gave people with blood cancer another chance at life and has since saved hundreds of thousands of lives. Her work with Stem Cells, Inc. involves the isolation of liver and neural stem cells as they pertain to a variety of diseases. 

Her most recent position was executive vice president for Scientific and Strategic Alliances at StemCells, Inc. During her 18-year tenure at StemCells, Dr. Tsukamoto led the scientific team that discovered the human central nervous system stem cell (HuCNS-SC®) and a second candidate stem cell for the liver and that transitioned the human neural stem cell into early clinical development in all three components of the CNS: brain, spinal cord, and eye. The biological potential and activity of these HuCNS-SC® cells were demonstrated in some patients and reflected results seen in preclinical rodents’ studies. The many challenges of developing a cell therapy in a small biotech firm led to the closure of StemCells, Inc., in August 2016.

She successfully invented the method to isolate blood stem cells in the body and obtained patent no. US5061620A

Source – [email protected]

This is the most valuable patent in her portfolio. Dr. Tsukamoto has 48 patents globally in her portfolio, which belong to 8 unique patent families.

#12. Laura Johanna van ‘t Veer

Dr. Laura Johanna van ‘t Veer is a Dutch Molecular Biologist and inventor of MammaPrint. Her research focuses on personalized medicine, to advance patient management based on knowledge of the genetic make-up of the tumor as well as the genetic make-up of the patient. She completed her Ph.D. in oncology and cancer biology from the Leiden University.

Laura is the Professor Laboratory Medicine and Director Applied Genomics Cancer Center at the UCSF (University of California San Francisco) since 2010. She has earlier worked with Agendia and the Netherlands Cancer Institute. Laura was also a Postdoctoral Fellow at Harvard Medical School (HMS) from 1989 to 1991.

Award & Honors:

  • European Society of Medical Oncology (ESMO) LifeTime Achievement Award, 2007 
  • Second prize EU Women Innovator Award, 2014
  • European Inventor Award in the category Small and Medium-sized Enterprises, 2015
  • European CanCer Organization Clinical Research Award, 2017
  • Precision Medicine World Conference Luminary Award, 2020
  • Recognized as one of the ’32 Amazing Women Inventors’, a group of women who succeeded in fields that are overwhelmingly dominated by men

Source – EPO

The most valuable patent in Laura Johanna van ‘t Veer’s portfolio is US7171311B2, for methods of assigning treatment to breast cancer patients. 

In Laura Johanna van ‘t Veer’s patent portfolio there are 40 patents globally, which belong to 8 unique patent families.

#13. Macinley Butson

Ms. Macinley Butson is an Australian inventor and holds a bachelor’s degree in science from the University of Wollongong. She is the founder of Passionately Curious, which provides access and opportunity to STEM, sparking curiosity for a generation of young minds. Prior to starting Passionately Curious, Ms. Butson has worked with Scilutions Pty Ltd as a Director.

She is notable as the youngest female inventor and scientist. She came up with her first invention at the age of 6. Ms. Butson has received numerous awards and honors as an inventor.

Awards & Honors:

  • Marie Claire + Bumble Glass Ceiling Awards, 2019 – The Future Shaper award winner
  • Australian Stockholm Junior Water Prize Winner, 2019
  • Instyle and Audi Women of Style Awards Judges Choice Winner, 2019 
  • Instyle and Audi Women of Style Next-Gen Innovator (Science) Award Winner, 2019 
  • Ozwater ’19 Keynote Speaker
  • [email protected] Speaker
  • AFR 100 Woman of Influence Finalist
  • 1st place Award at Intel International Science & Engineering Fair
  • 3rd place in Environmental Engineering at Intel International Science & Engineering Fair
  • NSW Young Australian of the Year, 2018
  • Event Speaker for [email protected]
  • 1st Place in Translational Medicine at Intel International Science and Engineering Fair
  • Australian Stockholm Junior Water Prize Winner
  • 4th Place in Energy: Physical at Intel International Science and Engineering Fair
  • 1st Place at the BHP Billiton Foundation Science and Engineering Awards

In Ms. Butson’s patent portfolio there are 6 patents globally and all the patents belong to unique patent families. She is an individual inventor of all 6 core patents. She invented an ultraviolet radiation sticker that measures the solar UV exposure required to sanitize drinking water, and a smart shield to protect women undergoing radiotherapy against excess radiation.

 Source – Good News Network

#14. Patricia Billings

Ms. Patricia Billings is an American inventor and businesswoman. She completed her study in Arts at Amarillo College in Texas. Her detour from art into technology came in the late 1970s, when a swan sculpture, after months of work, fell and shattered. Ms. Billings, who knew that Michelangelo and other Renaissance sculptors used a cement additive to give their plaster longevity, set out to create a modern equivalent.

Her specialty was plaster of Paris sculptures and Ms. Billings filed several patents for building materials including modular wall panels and roofing tiles.

The most valuable patent in Ms. Billings’s portfolio is US5647180A for a fire-resistant building panel marketed by the name Geobond®. 

Geobond® products are so resistant to heat that after being torched with a 2,000°F flame for four hours, it remains lukewarm.

Source – Inventricity

Ms. Billings in her patent portfolio has 8 patents globally, which belong to 5 unique patent families.

#15. Lynn Ann Conway

Ms. Lynn Ann Conway is an American inventor. After earning her BS and MSEE from Columbia University’s School of Engineering and Applied Science, Ms. Conway joined IBM Research. There she made foundational contributions to computer architecture, including the invention of multiple-out-of-order dynamic instruction scheduling. Fired by IBM as she underwent gender transition in 1968, Ms. Conway secretly started her career over again in ‘stealth mode’, soon becoming a computer architect at Memorex. She has also worked at MIT as a Vis. Assoc. Professor of EECS, Xerox Palo Alto Research Center, and DAPRA. At present, she is a Professor of Electrical Engineering and Computer Science at the University of Michigan.

Her specialties are computer science, systems architecture, electrical engineering, microelectronic design, research management, engineering education, human rights advocacy.

Awards & Honors:

  • Electronics 1981 Award for Achievement
  • Harold Pender Award of the Moore School, University of Pennsylvania
  • IEEE EAB Major Educational Innovation Award, 1984
  • Fellow of the IEEE, 1985, “for contributions to VLSI technology”
  • John Price Wetherill Medal of the Franklin Institute, with Carver Mead, 1985
  • Secretary of Defense Meritorious Civilian Service Award, 1985
  • Member of the National Academy of Engineering, 1989
  • National Achievement Award, Society of Women Engineers, 1990
  • Presidential Appointment to the United States Air Force Academy Board of Visitors, 1996
  • Honorary Doctorate, Trinity College, 1998
  • Electronic Design Hall of Fame, 2002
  • Engineer of the Year, National Organization of Gay and Lesbian Scientists and Technical Professionals, 2005
  • Computer Pioneer Award, IEEE Computer Society, 2009
  • Fellow Award, Computer History Museum, 2014
  • Honorary Doctorate, Illinois Institute of Technology, 2014
  • IEEE/RSE James Clerk Maxwell Medal, 2015
  • Honorary Doctorate, University of Victoria
  • Fellow Award, American Association for the Advancement of Science (AAAS), 2016
  • Honorary Doctorate and Commencement Address, University of Michigan, Ann Arbor, 2018
  • Pioneer in Tech Award, National Center for Women in Technology (NCWIT), 2019
  • Lifetime Achievement Award, IBM Corporation, 2020

Source – Michigan AI Lab – University of Michigan

The most valuable patent in Ms. Conway’s portfolio is US5652849A for an apparatus and method for remote control using a visual information stream. 

In Ms. Conway’s patent portfolio there are 5 patents globally, which belong to 5 unique patent families. 

Conclusion:

These are some of the female inventors from among numerous women who have contributed to the world of innovation. These women are an inspiration for young girls around the world. Women continue to disrupt the patent industry and make life easier with their inventions. We at PQAI salute and celebrate all the female inventors around the world. 

Patent Rejections: Do Not Lose Hope!

Patent Rejection - Don't lose hope

Got a patent rejection? Almost 90% of patent applications receive at least one non final patent rejections! Let’s dig deeper to understand the best practices while filing patent applications.

You have invested your retirement savings developing your idea and have spent years perfecting your invention.  After fine tuning your invention, you assist your patent professional with drafting your patent application, toiling to put together the perfect words to describe your invention.  Your patent drawings are detailed and meticulous.  After investing hours going over every word of your patent claims, you file your patent application.  Now, after years of waiting, you receive a notice from the Patent Office rejecting your invention.  Time to give up?  

No.  First, there are a number of resources that can assist you with understanding what the rejection is and how you might be able to overcome this rejection.  We will identify various resources that can assist you.  Reliance on these resources can be important since you might have a limited patent prosecution budget.  Also, it is likely that not only will you receive a first rejection (i.e., a first Office Action), but is also very likely that you will also receive a second rejection (i.e., a second Office Action) in the same application.  For example, according to a recent study by the US Patent Office, the average number of Office Actions that a patent examiner will prepare per issued utility patent where the patent finally issued in fiscal year 2019 was 1.7.  That is, on average, every patent that was issued in 2019 received at least one office action rejecting the patent claims.  

Second, once you receive a rejection of your invention, you are typically offered an opportunity to amend your application.  You can also argue against the Examiner’s rejections.  However, before you can make these types of arguments, you must have an understanding as to WHY the patent examiner has rejected your patent.  Such an understanding will allow you to engage the Patent Office in an effort to find some common ground in order for you to earn your patent.  Such an understanding will also allow you to seek assistance with overcoming this rejection and perhaps taking certain precautionary steps so as to avoid the same rejection pitfalls in the future.  Finally, we will identify certain precautionary steps that will assist you in overcoming these rejections, not only for the same patent application, but also for your future patent filings.  

Patent Rejections Are Common

Patent rejections are very common.  According to a recent study at Yale University, almost 90% of all patent applications receive some type of patent claim rejection.  So, you are not alone.  There is also hope since 60% of all patent applications will eventually (after an initial rejection) issue as a patent.  So, do not lose hope since the rejection that you received from the Patent Office is not a death knell to your invention.  Rather, a rejection is just the beginning of a process of negotiation with the patent office.  In this post we discuss both: precautionary measures you can take to avoid rejections as well as measures to take after receiving a (non-final) rejection.

Initial Precautionary Measures

There are a number of initial precautionary measures that you might want to consider before you file your patent application.  For example, one way to achieve a timely, cost effective resolution to seeking a patent is to hire a knowledgeable patent attorney before you begin to prepare your patent application.  A patent attorney will prepare your patent application for the rigors that it will experience during a patent examiner’s review.  Once the application is filed, the patent attorney will work with the Patent Office to marshal your invention forward so that your application will hopefully eventually mature as an issued patent.  Engaging a knowledgeable patent attorney will reduce the number of patent rejections that you receive, reducing the overall costs of patent “prosecution.”  In addition, your patent attorney will also hopefully expedite the process to eventually receiving an issued patent from the patent office.   

Precautionary Measures While Drafting Utility Patent Applications

The utility application accounts for the majority of applications that are filed in the US Patent Office and therefore will be the primary focus of the present article.  For example, according to recent statistics published by the US Patent Office, the Patent Office received roughly 621,000 utility applications in 2019.  (https://www.uspto.gov/web/offices/ac/ido/oeip/taf/us_stat.htm) 

Let’s understand the main components of utility patent applications and how can those be best presented in the application. Your utility application will generally include three principal components: a detailed written description, figures or drawings, and patent claims.  

Detailed Written Description

The detailed written description provides a list of all the elements of your invention.  The written description may include a description of how to make those elements, how to assemble those elements, and how to use the elements of the invention.  The written description explains what is illustrated in the figures and what the invention is as expressed by the words of the claims.  For example, the detailed explanation may include certain descriptors or headings such as the title of the invention, the technical area of the invention, background of the invention, a list of included drawings, and most importantly an in-depth, detailed description of your invention.  

One precautionary step that you might consider in order to avoid rejections based on your detailed description is to verify that all of the elements and components that you describe in your detailed description appear somewhere in your figures.  This also relates to any process or method steps.  Make sure each process or method step explained in your specification appears somewhere in one or more of your figures or drawings. 

Conversely, you should also confirm that all of the elements appearing in your figures are described or explained somewhere in our detailed description.  Also, when preparing the detailed specification understand that one purpose of the detailed description is that it should provide an understanding to a person of “ordinary” (not extraordinary) skill in the art on how to use and understand your invention.  So, if you think something should not be included in the detailed specification because you think everyone might already know how to perform a certain function, that is not the test.  If in doubt, include it in your detailed description.  This will help others understand your invention, and it may even help the Examiner when he or she reads your patent application.   

Drawings

The drawings should illustrate all of the components parts of your invention as explained in your detailed description and as recited in the claims.  Similarly, each element identified in your drawings should be mentioned in your detailed description.  Drawings should be included to understand your invention and must be included with your original patent application filing.  

One precautionary step that you might want to consider in order to avoid rejections based on your patent figures is to engage a professional draftsperson to prepare your drawings.  As they say, drawings may be worth a thousand words so it typically pays to get this portion of your patent application correct.  For patent drawings, like the words of your patent claims and like the words of your detailed specification, there are a set of rules and requirements that must be followed with your patent submission.  Those drafting professionals that prepare patent drawings are typically experts when it comes to these set of rules.  A true professional has a solid grasp on these patent regulatory norms but also a sufficient amount of artistic skills as professional patent illustrators and that can create patent drawings that are acceptable to the respective patent office.  Moreover, not only do they have the artistic skills, they are also typically well acquainted with computer design software.  Now, there are many types of software and computer systems that enable a professional draftsperson to create error-free patent drawings.  For utility applications, a typical draftsperson will charge about $30-50 per figure and for design patent drawings, about $40-60 per figure.  Of course, the complexity of your invention or the length of your application may require a higher charge.  

Patent Claims

The last few paragraphs of your application are the patent claims.  These are the most important part of your application because they define the legal boundary of your invention.  Your patent claims are different from your detailed description.  For example, your patent claims may be amended during your correspondence with the patent office.  In contrast, aside from amending typographical errors and the like, you cannot substantively amend your detailed description.  And, you cannot add new subject matter to your detailed description after you file your application.  

The patent claims of your patent application define exactly the limits of what your filed patent application does and does not attempt to cover (i.e., the “scope” of your invention).  As the owner of your patent, you have the right to exclude others from making, using or selling, only those items or things that are adequately described in your patent claims.  Because of the importance of the claims, your patent claims will receive the most amount of scrutiny from the patent Examiner.  

Filing Broad Patent Claims

The patent claims are prepared as the last few paragraphs of your patent application.  There are many strategic reasons to pursue broad patent claims when you initially file your patent application.  Naturally, the broader your patent claims, the broader the “scope” of your legal protection.  However, your broad claims must not cover what was known before you filed your patent application (i.e., the prior art).  Therefore, if you file broad claims, your broad claims will typically receive the most amount of rejections from your patent examiner. 

Filing Narrow Patent Claims

Alternatively, you may file narrow claims that try to capture only a narrow subset of technology.  For example, you may file a narrow “picture” claim that captures each and every technical aspect of a preferred design or application as it is illustrated in your patent application.  However, narrow patent claims could allow your competitors to avoid or “design around” your patent claims, giving competitors an opportunity to use your invention without technically violating your patent claims.  However, narrow claims typically will receive a lesser amount of rejections from your patent examiner.  As such, by submitting narrow claims, you may receive your patent earlier from the patent office.  

Patent Office Review

All three components of the Utility Application can receive “objections” or “rejections.”  As such, the Examiner can also take issue with your detailed description and/or the figures.  The key here to understand is that most, if not all patent application objections and/rejections can be overcome with changes made to your patent application detailed description, the figures, and/or the patent claims. 

Complete Application

One of the first things that the Patent Office will review is to see if you filed a complete application that includes all the necessary application parts.  After receiving an application, the Patent Office will review the patent application for completeness.  In order for your application to be complete and move on to the prosecution process where the Examiner will begin to scrutinize the actual technical substance of your patent application, your application must be complete.  

A complete application includes: 

  1. at least one patent claim, 
  2. an inventor’s oath or declaration, 
  3. payment of the required governmental filing fees, 
  4. and drawings, if necessary to convey the invention.  

The inventor’s oath or declaration is a document that identifies the inventors, and must contain certain required statements.  The oath or declaration must: 

  1. identify the inventor executing the oath or declaration by his or her legal name; 
  2. identify the patent application; 
  3. include a statement that the person executing the oath or declaration believes the named inventor to be the original inventor of a claimed invention in the application for which the oath or declaration is being submitted; and 
  4. state that the application was made or authorized to be made by the person executing the oath or declaration.  

If you file your application and your application papers do not include the inventor’s oath or declaration, or requisite fees, the Patent Office will send you a notice that you have submitted an “incomplete” application and that you are being required to provide the missing information.  If you receive one of these notices it is imperative that you respond and correct the deficiency noted by the patent office.  Typically, you are given two months from the mailing date of the notice to file the requested information with the patent office.  

Group Art Unit

After you file your complete application, your application will be assigned to one of the many Technology Centers at the Patent Office.  The Technology Center will review the application papers to determine what technical or group art unit the application should be assigned to.  Once the Technology Center identifies the proper technical subject matter of your invention, it will assign your patent application to a group of Examiners who have experience with the technology related to your invention. 

Patent Examiner Review

A typical patent application will undergo multiple levels of review by the Examiner during his or her substantive examination of an application.  First, the Examiner will determine whether to reject your application and issue a restriction requirement.  Second, the Examiner will review your application and analyze whether one of the primarily four types of rejections are applicable.  

Restriction Requirement

For example, an initial examination next step in the patent application review process is for the Examiner to review your application for a potential restriction requirement.  The most common type of restriction is when you attempt to claim multiple inventions that are either independent (i.e., completely unrelated) or distinct (i.e., related but capable of separate manufacture and separate use).  The Examiner will determine if your patent claims are trying to capture more than one invention.  The theory here is that examination would be too time consuming and burdensome for the Examiner to search for more than one invention when the fees that you paid when you filed your application only paid for the Patent Office to perform a single prior art search, not multiple searches for multiple inventions.  

Example Claims

For example, let’s say you submit a patent application that is directed to a new idea for making orange juice. Your patent application includes a number of patent claims.  Specifically, your patent application includes a first patent claim that is generally directed to your orange juice making machine and includes the following elements:

1.  a mechanism for selecting an orange, 

2.  a mechanism for squeezing the orange, 

3.  a collection device for collecting the orange juice, and 

4. a disposing device for disposing the now empty orange peel.  

Your patent claims also include a claim that is generally directed to your method of making the orange juice and includes the following four steps: 

1. selecting an orange, 

2. squeezing the selected orange, 

3. collecting the juice from the squeezed orange, and 

4.  disposing of the empty orange peel.  

Claim Election

As you have presented a machine patent claim and a method patent claim, the Examiner may issue a restriction requirement and thereby reject your claims.  Such a rejection may require that you “restrict” your proposed patent claims to proceed with selecting either your orange juice machine related claim or your orange juice making method claims.  

By selecting one claim over the other claim, you are not sacrificing the unselected claim.  For example, if you opt to pursue your orange juice making machine of claim 1, you will not be sacrificing your method claims.  You can file what is called a “divisional” patent application to seek protection of your unselected method claim set.

Substantive Examination

Once you have successfully narrowed your application down to a single invention, the Examiner will move onto the substantive examination phase.  During the examination phase, the Examiner may issue formality rejections and these are generally related to Sections 101 and 112 of the Patent Act.  For example, as noted in the graph provided below, the Section 101 and 112 type formality rejections account for almost 30% of all rejections, with Section 101 rejections accounting for only about 8% of all Examiner rejections.

For rejections based on the prior art, there are two primary rejections: a novelty rejection under Section 102 and an obviousness rejection under Section 103, both under the Patent Act.  Again, as noted in the graph provided above, rejections based on Section 102 account for roughly 21% of all rejections whereas rejections based on Section 103 account for roughly 47% of all rejections.  

So, let’s take a quick look at each of these rejections and why the Examiner might issue such a rejection.  After discussing the substance of the rejection, we will identify certain precautionary measures you might use to avoid receiving these types of rejections in your current patent application as well as potential future patent applications that you might file.  

Section 101 Rejections

Rejections according to Section 101 relate to the idea of “double patenting.”  A patent is a government grant that gives you the exclusive rights to your invention, for a limited period.  According to Section 101, you should not be allowed to extend the time limit by obtaining multiple patents for the same invention or for obvious variations of your invention.  

There are two kinds of double patenting: one based on the law and the other based on a question of obviousness.  First, there is “statutory” double patenting which prevents you from having two patents with the same invention.  The prohibition against statutory double patenting arises from the Patent Act, which states that an inventor can “obtain a patent.”  Because double patenting requires the same claim in two patents, a statutory double patenting rejection is relatively easy to avoid and is fairly uncommon. 

The second type of Section 101 rejection occurs much more frequently.  “Obviousness-type” double patenting prevents you from obtaining a patent with a claim that is obvious over a claim in another of your patents.  The principle behind the doctrine is that you should not be able to extend the life of your first patent by obtaining a second patent with a claim to an obvious variation of your previous invention.  

One way to overcome these “double patenting” rejections is to file a legal document called a “terminal disclaimer.”  In the disclaimer, you agree that your second patent will expire at the same time as your first patent.  The Patent Office provides a standard form for filing such a terminal disclaimer.  A copy of this form can be found here

Patentable Subject Matter 

When the Examiner reviews your patent application, the Examiner will ensure that your application meets the legal requirements for patentability.  This includes whether your patent application is the right subject matter for protection under Section 101 of the Patent Act.  Section 101 states that patents may be granted on “any new and useful process, machine, manufacture, or composition of matter.”  If your invention is not a process, machine, manufacture, or composition of matter, your invention is not patentable because your invention is not the right type of invention and will therefore be rejected.

While most inventions fall within one of these categories, some inventions do not.  For instance, if your patent application claims that your invention is an electromagnetic wave having certain characteristics, your application will be rejected because a wave is not a “process, a machine, a manufacture, or a composition of matter.”  Pure data (meaning data claimed alone as opposed to data residing on a computer or data being manipulated during a computer process) is also unpatentable and will be rejected.  

If your invention is going to be patentable, your invention must also avoid falling under a judicially created “exception” to patentable subject matter.  While the articulation of these exceptions has varied over the years, what is clear is that abstract ideas, laws of nature, and natural phenomenon are exceptions to statutory subject matter.  Inventions that are considered abstract ideas or laws of nature are not patentable because they are not the right type of invention.  

These exceptions to subject matter eligibility have become much more important in the last few years, especially for software inventions.  In fact, according to a recent post at ipWatchdog, approximately 60% of software patent applications reviewed by the Patent Office are initially rejected under Section 101 because the software invention is considered to be directed toward an abstract idea. (https://www.ipwatchdog.com/2019/08/13/update-101-rejections-uspto-prospects-computer-related-applications-continue-improve-post-guidance/id=112132/)

If your application has received a Section 101 rejection, that means the Examiner believes that your claims relate to a type of invention that is ineligible for patent protection.  There are really two different types of Section 101 rejections.  First, either the Examiner is saying that your invention does not fall within one of the four statutory categories.  Or second, the Examiner is saying that your invention falls under one of the patent ineligible exceptions (which is usually either an abstract idea or a natural law/phenomenon).  In either case, such a rejection does not have to be the end of your patent application, although the rejection can sometimes be difficult to overcome a Section 101 rejection.

Section 101 Precautionary Steps

As a precautionary measure, you may consider reviewing your claims before you file with the patent office with an eye towards a potential 101 rejection.  In most cases, these types of 101 rejections can be overcome by revising your application to explicitly claim one of the four statutory classes: a “process, a machine, a manufacture, or a composition of matter.”  For instance, instead of claiming an electromagnetic wave having certain characteristics (which is a non patentable invention), you can revise your claims to cover a method for creating that wave (statutory under the “process” category).  Alternatively, you can revise your claims to cover a device that creates the wave (statutory under the “machine” category).

In some cases, the broadest reasonable interpretation of a claim may cover both statutory subject matter and subject matter that does not fall within one of the four classes.  For instance, claims to “machine-readable media” include both physical computer memory and transitory electromagnetic waves.  These claims should be revised to remove the possibility of the claims covering subject matter that falls outside the four statutory classes.

Section 102 – Novelty of Invention

Prior art rejections are the most common rejections issued by US Patent Examiners.  For example, as noted in the graph reproduced below, almost 70% of all rejections issued by Examiners are based on some prior art that is used to reject patent claims.  

In order to obtain a patent on your invention, one of the legal requirements is that your claimed invention (i.e., the one defined by your patent claims) must be “new” or “novel.”  In other words, you must be the first inventor to have invented your invention before anyone else.  So, there can be no prior invention before you filed your patent application.  A prior invention is typically something that was publicly known or publicly available before your invention.  

These are very common rejections.  Indeed, roughly 21% of all Examiner rejections are related to novelty rejections under Section 102 of the Patent Act.  So, one of the more common rejections is that the Examiner will say that your patent claim is not novel or is not new because someone somewhere was the first to practice or teach your invention.  They were the first to invent your invention.  You were the second to come up with this invention.  Therefore, your patent claims will be rejected.  

For example, returning to our sample claim, let’s assume that you have decided to pursue your orange juice making machine invention.  The Examiner will review your claim and then search through the prior art in effort to determine if he or she can find the following in the prior art: 

An orange squeezing apparatus that includes 

1. a mechanism for selecting an orange, 

2. a mechanism for squeezing the orange, 

3. a collection device for collecting the orange juice, and 

4. a disposing device for disposing the now empty orange peel.  

The Examiner will search available sources (e.g., granted patents, scholarly publications, industry publications, etc.) in order to ensure that your application’s claimed orange squeezing invention deserves patent protection.  The Examiner may also use non-patent literature including magazine articles, newspaper articles, electronic publications, on-line databases, website, or internet publications.  If the Examiner identifies prior art that he or she believes shows an orange juice making machine including all four machine elements that are recited above, your claims will be rejected as lacking “novelty.”  Basically, the Examiner is saying that your invention is not “new” or “novel.”

One Reference Showing All Elements

Patent novelty, therefore, refers to the uniqueness of your invention.  Your invention will be novel if no single prior art reference discloses all the elements that form your claimed invention.  So, there are two critical pieces of information that you must analyze to determine novelty.  First, you must identify the expressed words of your patent claims and all its limitations.  Second, you must review the cited printed publication that the Examiner relies upon and understand what the printed publication teaches.  An application may be rejected under Section 102 only if a single prior art reference matches each and every element of your claimed invention.

Another important issue is that the printed information that the Patent Examiner is relying upon must be “prior” to your invention.  That is, the printed information must bear a publication date that is earlier or “prior“ to your filing date.  If the date of the cited information is not earlier than your filing date, the cited information is not “prior art” and therefore must be removed from the Examiner’s analysis.  

Here is an example of a hypothetical anticipation rejection.  Recall our hypothetical invention for your orange juice machine.  As noted, your orange juice machine invention includes four components: 1. a mechanism for selecting an orange, 2. a mechanism for squeezing the orange, 3. a collection device for collecting the orange juice, and 4. a disposing device for disposing of the empty orange peel.  

Now let’s say the Examiner rejects your orange juice machine claim.  Specifically, the Examiner states that your orange juice machine is not patentable as your invention is lacking novelty over a patent published in the United Kingdom and that names John Smith as an inventor (i.e., “the Smith patent”).  The Smith patent, written in English, is entitled “Orange Juice Making Machine.”  Upon a review of the Smith patent, you recognize that, yes the Smith patent is generally directed to an orange juice machine.  However, you recognize that the Smith patent does not teach any type of mechanism that allows for the selection of one orange over another orange prior to the orange being squeezed.  

As you describe in your detailed explanation of your patent application, one of the advantages of your invention is that the selection of an orange is based on the size, the color, or some other characteristic of the orange.  Your invention allows your apparatus to select a higher quality orange over a lesser quality orange.  In contrast, you note that the Smith patent merely describes an orange juice maker that randomly selects oranges, irrespective of the color, the size, or some other characteristic of the orange.  In the Smith patent, there is simply no “selection of an orange.”  And since your patent claim expressly recites “a mechanism for selecting an orange,” the Smith patent does not anticipate because Smith does to disclose at least the noted important limitation of your patent claims.  

Since anticipation requires that all of the elements in your claim must be found in a single prior art printed document, you can argue to the Examiner that the Smith patent does not “anticipate” because Smith lacks a teaching of some type of mechanism that selects an orange. 

Another option to address the Examiner’s anticipation rejection is to amend or clarify your orange juice apparatus patent claim.  For example, you can amend your orange juice machine claim to clarify an element, add a new element, or do both.  You must, however, have support for your proposed clarification in your application as filed.  In our example, if the Smith patent fails to disclose “a mechanism for selecting an orange,” you can amend your claim to further differentiate your invention from the orange juice machine disclosed in the Smith patent.

As one example, you could amend your “selection mechanism” claim limitation to now read: “a mechanism for selecting an orange based on a characteristic of the orange.”  (claim clarification shown in underline).  The amendment would further distinguish your claimed invention from that taught by Smith’s patent.  

Reference Must Be Prior Invention

Another important issue is that the printed information that an Examiner is relying upon as “prior” art must bear a publication date that is “prior to” or “before” your invention.  If the date of the cited information is not earlier than your filing date, the information is not a “prior” invention since the information was published after your invention.  And since the cited information is not “prior art,” the cited information cannot be considered by the Examiner since the information cannot anticipate your invention since the information was not “before.”  

Returning to our hypothetical rejection of your orange juice machine, let’s say that upon a closer review of the Smith patent, you observe that the patent was originally filed in the UK.  As is typical with UK or European type documents, you note that the Smith patent uses a European date format.  That is, the Smith patent references a publication date by way of a date format where the day comes first, then the month is referenced, and then the year is listed.  The Smith patent identifies a publication date as:  “01/10/2020.”  Well, this UK styled publication date means that the publication date of the Smith patent is the first (1) day in October (the tenth month) in the year 2020.  The date does not mean that the publication date of the Smith patent is January (the first month) 10, 2020.  

However, you note that the Examiner’s rejection states that the Smith patent is “prior art” to your invention because the Smith patent was published before your patent application.  But your patent application was filed on April 1, 2020.  A proper reading of the Smith patent would show that Smith is not “prior” to your invention since you filed for your invention a full six months before the Smith patent was published on October 1, 2020, and not January 10, 2020.  Therefore, the Smith patent must be removed as an “anticipating” reference since Smith is not prior art to your patent application filing.  

Section 102 – Precautionary Steps

If you do not have a budget to hire a patent attorney, another way to achieve a timely, and cost effective resolution to seeking a patent is to engage a patent professional searching company to perform prior art search.  There are many prior art search firms and companies that perform these types of searches.  A typical prior art search typically costs on the order of about $300-500 for a typical mechanical invention.  There are a number of benefits to performing a precursory prior art search, one conducted before expending the time and resources to prepare and submit a patent application for review by the Patent Office.  For example, the benefits of conducting a preliminary prior art search include: 

1. Avoid submitting patent applications with claims that are not patentable and will be easily rejected.

2. Determine whether your invention is novel compared to public prior art. 

3. Develop a strong patent claim strategy before you file your patent application (and reduce the chance of extensive amendments). 

4. Account for close prior art when drafting your patent application.  For example, you might want to describe advantages or improvements over relevant prior art, as this can help persuade the patent office that your invention is “non-obvious.”

5. Understand how your idea fits into the technological field.

6. Be better prepared to discuss your invention with a patent examiner and explain what aspects of your work might be patentable.

Basically, a patent search will give you an idea about whether it even makes sense to pursue a patent in the first place.  Unfortunately, patent searches do not come with guarantees.  However, the goal of a patent search is to reach perhaps a 75-80% level of confidence threshold.  To reach a higher confidence level would take thousands of dollars, and to reach near certainty would require many thousands of dollars, so the search that is undertaken is reasonable given the value of the invention.  It is also reasonable given that the prior art represented in patent applications filed for the first time within the last 18 months are simply not searchable because they are required by law to be kept secret.  So a “no stone unturned” search is not possible and not economically wise.  But a thorough search of what can be reasonably found leads to better decisions and always leads to a better written patent application that takes into account the prior art.  Without knowing what is in the prior art, there is simply no way to accentuate what is most likely unique in comparison to the prior art.  In other words, without a search you are describing your invention in a vacuum.  Indeed, returning to our hypothetic rejection over the Smith reference, perhaps a basic search would have identified this reference.  With this knowledge in hand, perhaps you would have crafted your patent claims to already include a patent claim limitation that is not taught by this reference, saving you both time and money.  

If you do not have a budget to fund a third party prior art search, you can perform your own patent search which will allow you to gain a better understanding of the prior art.  For example, there are multiple resources that can assist you with this basic search including at least the following: Google Patents, USPTO search interface, Espace (European Patent Office), WIPO search interface etc.  There are a few more free search tools available.  Performing such a search will allow you to prepare and submit a higher quality patent application since you now know what the patent Examiner may cite against your patent claims.  This will also allow you to familiarize yourself with certain prior art that the Examiner may use in rejecting your claims.  Performing a prior art search is yet another method that will reduce the number of patent rejections that you receive, reducing the overall costs of patent “prosecution”.

Another cost effective pre-filing resource for low budget patent filings is Project PQAI.  Project PQAI is a collaborative initiative to build a common AI-based prior art search tool.  PQAI stands for Patent Quality through Artificial Intelligence.  One of the advantages of PQAI is that it attempts to make the patent process more transparent for all involved, and to significantly grow the number of inventors, thereby accelerating the pace of innovation while simultaneously improving the overall quality of patents.  By using PQAI and its AI based searching algorithms, cost free searches of hundreds and thousands of potentially relevant references can be reduced to a best of prior-art collection in a top-10 results format.  By providing you with a “top-10” search result format, this will save you the aggravation, time, and expense of refining and detailing your search queries over and over in an attempt to find the best prior art references.  

Moreover, with PQAI, you are provided with a query mapping table.  With such a table, you are provided with a first column that shows a part of the invention query and a second column that shows the relevant text from the search result.  This mapping is not just word-to-word but it is highly contextual, unlike many other search results provided by third party search firms.  

Best part is – the inventors who are not skilled with core patent search skills can also perform prior art search using PQAI themselves.

Section 103 – Obviousness of Invention

A rejection under Section 103 of the Patent Act is the most often type of rejection that an Examiner will issue.

Combination of References

An Examiner may reject your patent claim based on Section 103 when your claimed invention is not identically disclosed so the reference teachings must somehow be modified in order to meet the claims. 

Recall from our example that the Smith patent describes an orange juice maker that randomly selects oranges, irrespective of the color, the size, or some other characteristic of the orange.  We already noted that in the Smith patent, there is simply no “selection of an orange.”  And since your patent claim expressly recites “a mechanism for selecting an orange,” the Smith patent does not anticipate because Smith simply fails to disclose the important claim limitation.  

In response to your claim modification, lets say that the Examiner issues a second Office Action and cites the Smith patent in combination with a second patent: the Jones patent.  Specifically, in the Office Action, the Examiner notes that the Smith patent might not expressly teach a “selection mechanism,” but the Jones patent makes up for the missing element.  Now, the Examiner states that the combination of these two patents teaches all of the elements of your orange juice making machine and therefore issues an “obviousness rejection” of your claims.

In our hypothetical, you note that the Jones patent is generally directed to a method for making wine from squeezing grapes.  In the Jones patent, a device is used to select a grape based on a characteristic of a grape.  The Examiner’s position, therefore, is that a person having ordinary skill in the art would start with the teachings of the Smith patent and then make the modification as suggested by the Jones patent.  Therefore, the Examiner reasons, your juice making machine patent claim is “obvious” over the combination of the Smith patent and the Jones patent.   

Amend to Include Additional Claim Limitation

There are many strategies that you can use to respond to the Examiner’s “obviousness” rejection.  A few of the more frequently used strategies are mentioned below. 

First, if the combination of prior art really does show all the elements of a particular claim, you can amend your claim to clarify a feature or to add an element.  In the above example, you can add a further element to the limitation of “a mechanism for squeezing the orange” and further specify the structure of the mechanism like, “a robotic hand for squeezing the orange.”  In conjunction with the claim amendment, you could argue that the combined Smith and Jones patents altogether fail to show the new claimed combination.  Of course, the Examiner may search the prior art further and find a third reference that discloses “a robotic handle.”

No Motivation to Combine References

Alternatively, you could argue that there would be no motivation to combine the Smith and Jones patents together as suggested by the Examiner.  For example, if the Smith patent expressly states that certain features should be avoided (like the use of selection mechanism), there would be no motivation to combine the primary Smith patent with a secondary Jones Patent disclosing the very feature to be avoided.  One way to argue against the motivation to combine is to study the intended purpose of each prior art reference.  Would the Examiner’s combination go against the intended purpose of one or more prior art references?

The Examiner’s reliance on 103 rejections has been slightly increasing over the last two-three years.  For example, the graph provided below shows the percentage trend of 103 rejections issued by patent examiners: 

Macintosh HD:Users:thomaswettermann:Desktop:Screen Shot 2021-01-28 at 8.39.06 AM.png

Section 103 – Precautionary Steps

As noted in the graph provided above, there is almost a 50% chance that the rejections you receive for your patent application is related to Section 103 of the Patent Act.  As such, you should consider utilizing PQAI which will help you identify the combinational prior-art that could be cited as the basis for a 103 type rejection.  PQAI has an open-source search engine that has the capability to identify the combinational prior art.  For example, if we were to input our hypothetical “orange juice” making machine as an idea query into PQAI, we would receive a first combination of results, which may have identified the Smith and Jones references, as described above.  As such, running your original “orange juice” making machine thought PQAI once to look for potentially relevant combinational prior art thereby allowing you to identify potential roadblocks to you seeking a patent.  Once these potential roadblocks are identified, you can draft your patent claims accordingly to avoid such potential 103 rejections.  

Conclusion

When faced with a rejection of your invention, you must try to understand the basis for the rejections and what is required to overcome these rejections.  You should not lose hope since changes can be made to address the typical rejections raised by the Examiner.  

Exclusive Interview with Sam Zellner | Inventor Spotlight

Exclusive Interview with Sam Zellner

Sam Zellner is a Prolific Inventor, an Entrepreneur, Adept Portfolio Manager, Product Lead for PQAI and Ex Director Innovation at At&T.

Mr A: Core innovation happens when we stop believing in the societal norms of accepting ‘that’s how it works’

Sam Zellner: “The corollary to this is believing long held assumptions can’t change. Ken Olsen, CEO of Digital Equipment Corporation (DEC) said in 1977, ‘there is no reason anyone would want a computer in their home’. The challenge for all of us is realizing when a basic assumption is no longer true. The funny thing is it’s always obvious later on!”.

Introduction

Samuel N Zellner is a prolific American inventor with more than 200 issued and pending patents worldwide. He has held many prestigious positions in the IP fraternity. Sam Zellner retired as Executive Director, Innovation at AT&T in January, 2010. During his tenure at AT&T he created state of the art platforms utilizing Artificial Intelligence (AI). He also developed new approaches in advanced big data concepts to develop high-value patent portfolios and monetize these Intellectual Property (IP) assets. 

His current projects involve creating an open source combinational prior art search engine utilizing AI called PQAI, which stands for Patent Quality through Artificial Intelligence. Sam Zellner is also working on InspireIP, an invention disclosure system making invention management easy for inventors and IP counsels.

Sam Zellner is a board member on a number of the state IP Alliances as well as the newly formed  US Intellectual Property Alliance.  He has also been recognised as IAM 300 Top Strategist 2019. He is experienced in planning and strategizing in high tech. Sam is also on the board of directors of the Licensing Executive Society (LES), Atlanta chapter since January, 2018. 

Exclusive Interview With Sam Zellner

We asked some questions from Sam Zellner and through his experience, he has provided some brilliant insights for the inventors and patent portfolio managers.

#1. What challenges do you face in your daily life as an inventor?

My big joke is that the inventors are some of the loneliest people as they are not able to find support for their ideas. As an inventor, you come up with an idea and if you present your idea to someone, they tend to discount it as being bizarre or incorrect.  

For example, way back when the inventors were thinking about putting cameras on cell phones, everybody was like why would we put cameras on cell phones? data transmission is  expensive, cell phones cannot hold much data and At the time it seemed like a crazy idea.It is hard as an inventor to share  ideas with other people, as inventors typically base their inventions on assumption sets different from the accepted norms. Battery technology will improve (think electric cars), people will change their behavior (think buying online), laws will change (think Uber and taxi licenses).  This is why most people can’t see or accept inventors’ visions.  Later on, when hopefully the idea is adopted, everyone’s lense looks at the concept with the new assumption set will say that either ‘I was also thinking of this idea back then’ or that ‘it was obvious’. It is really hard as an inventor to get credit. With the patent system, the inventor gets some credit as they are recognised with the patent. 

Generally, it’s sort of a lonely life as an inventor, a tough life because very rarely do people acknowledge that you had a good idea. Rarely do you get recognised as doing something novel, rather you are recognised as crazy, which is the common thought process.

#2. Are you part of any inventor groups or community?

I am not aware of a lot of communities. Maybe, the individual inventors are a part of some communities. My experience with corporate inventors is that they tend to talk to their associates, but I am not familiar with the corporate inventors being part of a specific group. You might check with the inventors’ association to see if there are any particular groups that they are pushing towards. I think in Atlanta there are some incubators that have events, which are fairly popular. Tech Village in Atlanta is one.

#3. What motivates you to invent?

As a lot of people say, engineers like solving problems, I think it’s a mixture of curiosity and wanting to solve problems. Patents are about solving problems, so it comes naturally that way. I look at problems and try to think of how to solve them.

#4. When do the best ideas occur to you?

I think most people say that when you are in the shower. On the contrary, I think typically, like I said it’s about solving problems, so the best ideas come a few hours after you see a problem or run into a problem. As they say, your mind is thinking about a problem and to invent, sometimes you really need, almost, the subconscious to be helping you. Because unfortunately, the assumptions that most of us go around with are so strong that it’s really hard to see past those assumptions. Particularly, what are called the ‘old assumptions’ – assumptions that might have been good a year ago but now because of new technologies, change in economic factors or regulatory landscape or something in the environment. Now the old assumption that – ‘we can’t do this’, is probably no longer valid. Then all of a sudden, when you think of assumptions as walls and when you move that wall away, then a whole gamut of opportunities open up. 

It reminds me of location services, I did a lot of patents around location services. Before, we had no real in-location services capability, GPS came and then we had the enhancements with cellular, allowing us to do location for 911. The general thinking of the people was that ‘I don’t know where somebody is when they call, when they are using a phone’, now all of a sudden, anybody can know exactly where the other person is. As an inventor, now you wonder as to what you can do with that. What came to my mind, at the time when it started off, was that the cellular network moves when I leave the house. The cellular network knows when I leave the house, it can see me driving my car, location is changing, then it should know how to change the thermostat in my house. So it’s more efficient, I am saving energy because it used to always bother me that when I leave the house, since in Atlanta it’s very hot, the air conditioning is always on and wasting power that way. So it could let the house get a little warmer when I am not there, there’s no harm and I am saving money, energy. And ideally the cellular network could see when I am coming back again so it could turn the air conditioning on and when I get home, the house would be cool again. So it’s a simple example, once it’s realised that location can be used to control things, now it opens up all kinds of opportunities.

Another thought was now that the cellular network can see me arriving in a city, if there’s a hotel in that city, it can automatically register me for the hotel because it can see that I am coming for my hotel room. So it’s amazing, once I start realising and accepting the fact that I can know where people are, now I can make some assumptions about what should be done based on where I am. So that’s what I mean about the whole idea of changing assumptions and opening up more opportunities.

#5. Is there a systematic approach to coming up with innovations? 

There are a lot of techniques out there, it depends on the person what technique works well. Everybody is different in how they think so it can be different. There are really two parts to this, one there’s getting the seed idea, identifying the problem and on the cusp of solving the problem and then there’s also sort of extending it. So the hard part is getting the seed idea and finding a problem that’s of significance, which hasn’t been solved well. For me, I go through the assumptions that I am making about the problem and test each one to see if it’s true. If I take this assumption, what it does. And that for me helps quite a bit. 

The other piece, which I see a lot of people do and is easy to do, is that as in my example before, about location services and changing the thermostat, people tend to get fixated on one solution to a problem. They don’t really generalise it because again, think about it, the patent is looking into the future. As an inventor you are trying to throw a solution out into the future and it’s very hard to know how the world will change in the future. Therefore, you want to expand your idea plus you don’t know what people have really done. 

That’s one of the reasons why PQAI is so helpful. When you run PQAI, you can see where the thinking is and you can modify your idea based on that. As an inventor, you might find out that people have already thought about your idea, so you might want to think about the next generation of the idea. Maybe there are some aspects of using location to control something, what would come next and where else might you apply, if it’s just thermostats. What about using location to provide package delivery notifications and you know there’s lots of other things there.

When you get a seed idea for an invention, try to generalise it. I think about it as trying to generalise it till all of sudden it’s no longer novel, it gets so broad that you’ll run into the wall that says, it sounds familiar or that’s already been done. In any case it helps you, particularly in the patent because as you know, with a patent you have your initial claim and then you have your dependent claims. Thus, it expands your idea and this way if you do a mapping, like again in the example of location services and controlling the thermostat, you might generalise it from controlling a thermostat to controlling a device and you might define control as turning on and off instead you might want to define control as adjusting or you can say controlling multiple devices. So that helps to broaden out your idea in case some prior art is already there, you can find your segment.

#6. What was your first invention and when?

First invention… I didn’t go anywhere with that but I tried. I had two ideas, one idea was in 1983, creating a phone ringer that would play tunes. My interest was particularly in the fight songs in colleges I attended school at Northwestern University. They have a big marching band and they have their own fight song like most universities do. And I thought that wow all the alumni would love to have their phones ring their school’s fight song. So that was my idea and this is prior to cell phones, so this is at the time of landline phones. At the time, the landline phones did not have any tunes playing, they were basically just a standard ringer. So I was trying to put together the electronics around it and unfortunately, I could not quite get it together. 

The other idea, which sounded crazy back then, was putting TVs in an elevator. I used to work in a high rise building back then. I just noticed how much time people spend in the elevator and how uncomfortable people were in the elevator. Then, I thought to myself, wow, if you could put a TV in there and show some news or something, that would be actually welcome, since people are looking around uncomfortably in the elevator. I actually talked to the city of Chicago elevator commission about putting TVs in the elevators and they thought I was crazy.  Now the ironic thing is, I haven’t seen that many but there are a few TVs in the elevators but you see TVs in public places. It is one of the examples where I should have been thinking broadly because now you see TVs at the airport, gas stations, pumps, and in a lot of different places. It goes back to thinking broadly because sometimes your initial use case is not the most important use case.

#7. What shall be your advice for budding inventors?

Run your idea through PQAI and gather some knowledge about how other inventors have tried to solve the problem. Keep an open mind. Don’t get totally stuck on your one use case. Listen to people, share your idea, obviously in a way that it is protected but maybe after your provisional application or with your close friends to try and get a sense of how people are reacting to your idea. A lot of times it will give you clues as to maybe where you are a little off in your idea. It’s very rare, in my experience, that people have hit it right in their initial idea. They are in the right area, they have the right basic building blocks but it needs to be adjusted in some way. My advice is be open and listen carefully to people’s reactions as it might give you clues for where you should be going.

The other thing is that inventing is very hard. Don’t be discouraged if your first idea might not be novel. It is very hard, you are competing against all the inventors in the world. That’s very tough so don’t get discouraged.

#8. How was your experience as an inventor at AT&T?

My experience at AT&T was very good. AT&T has a very energetic, creative environment and very smart people. We could talk about the new ideas and people were very open to it. We were working with a lot of cutting edge technologies at AT&T. So I found it very easy to come up with new ideas in that environment.

#9. What are some tips you would like to give to a patent portfolio manager?

Again, to have an effective patent one really needs to be broad. So I would want to encourage the patent portfolio managers to make sure the patent is broad enough so that as the future unfolds, the patent is still relevant. I think what helps to broaden your patent out and obviously, to test it is to do some prior art searching. The prior art searching really gives you a sense of how other people are thinking about the idea. Then you can see how your idea relates to those thoughts, that usually generates more use cases and more thoughts about how to broaden out the patent and where novelty really exists. 

I would encourage the patent portfolio managers to do some prior art searching and that’s where PQAI provides great opportunity as prior art search takes a lot of time. With PQAI you can do it very quickly. It is sort of a golden opportunity for patent portfolio managers to leverage it and ensure that either the ideas/inventions are new disclosures or continuations or their very best.

Sam Zellner | Patent Portfolio

The statistics and charts hereunder provide an insight into Sam’s patent portfolio, which has more than 200 issued and pending patents worldwide. 

Note – Patent families represent the count of total unique patent families. Patents represent the total number of records i.e. considering all the family members of an INPADOC (International Patent Documentation) family. The following four statistics are based on unique families count.

  • Technology Area And Patent Families Count –

Sam’s patent portfolio has 292 patents globally, which belong to 91 unique patent families. He has worked in many areas of the tech industry but mainly, most of his inventions are related to electronic communication techniques and instruments. The count of inventions in this and related domains is 86.

The chart below details the areas of technology in which his patents have been filed:

Sam Zellner Patent Portfolio
  • Technology Through The Years – 

This statistic is based on Sam Zellner’s patent filings periodically, indicating as to how many patents are filed year by year and in which area of technology:

Sam Zellner Patent Portfolio
  • Inventions – 

Sam Zellner has patents in 91 different patent families within his patent portfolio. He is an individual inventor of 17 and a co-inventor of the rest of the 74 core patents:

Sam Zellner Patent Portfolio
  • Patent Assignment 

Sam Zellner is affiliated with AT&T Inc, putting AT&T on the top of the list of patent assignments by Sam for his inventions. However, there are a few more names of other assignees in the list, in the cases where Sam’s inventions have been re-assigned by AT&T and all these patents were filed by AT&T. All the subdivisions of AT&T as AT&T Inc have been considered.

Sam Zellner Patent Portfolio

The term “Patent Counts” represents the counts of individual patents filed in various countries, irrespective of the patent family. The following statistic is based on the total number of patents in the portfolio: 

  • Patent Filing Worldwide

The following graph shows Sam Zellner’s patent filing for inventions worldwide. Majority of the patents have been filed in the United States of America. Also, there are 10 patents, in which the applications have been filed before the World Intellectual Property Organization (WIPO) & the European Patent Office (EPO).

Sam Zellner Patent Portfolio

The Key Takeaways

Sam Zellner has decades of experience as an inventor and a patent portfolio manager. Sam’s advice to the inventors is to think broadly and expand the idea beyond your first use case. Sam Zellner recommends the inventors to keep an open mind and observe how the people are reacting to the idea, when shared with people in a protected way and also motivates them to not get discouraged in case their idea is not novel. He encourages the patent portfolio managers as well as the inventors to do some prior art searching, to find how other inventors are approaching the same problem.