Patent Text Research Articles

Patent Text and Long-Run Innovation Dynamics: The Critical Role of Model Selection

Patent Text and Long-Run Innovation Dynamics: The Critical Role of Model Selection

Knowledge Sources, Novelty, and Generality: Do Standard-Essential Patents Differ From Nonstandard-Essential Ones?

This article comprehensively investigates the differences between standard-essential patents (SEPs) and nonstandard-essential patents (N-SEPs) in terms of knowledge sources, novelty, and generality and investigates the impact of knowledge sources on novelty and generality. Specifically, we take a sample of declared SEPs related to long-term evolution standards and their corresponding N-SEPs from 2006 to 2015. To explore the differences between SEPs and N-SEPs in terms of knowledge sources, we used patent bibliometric analysis and text analysis methods to mine the technological domain, scientific literature, and standard literature involved in the backward citations of patents. Then, three indicators are established: 1) technological diversity, 2) science linkage, and 3) standard linkage. Standard linkage has rarely been discussed in the literature. We used text analysis to construct two indicators, the novelty and generality indices, which depict the differences in novelty and generality between SEPs and N-SEPs by comparing the text similarity of focal patents to prior and subsequent patents. The impact of different knowledge sources on SEPs and N-SEPs was verified through empirical analysis. This article innovatively found that SEPs are less technologically diverse and scientifically related, whereas their standard linkage is higher than that of N-SEPs. SEPs are more generic and less novel than N-SEPs. In addition, different knowledge sources have distinct effects on the novelty and generality of SEPs and N-SEPs. These findings can provide a reference for firms’ knowledge acquisition, resource allocation, and selection of R&D partners in the patent creation process.

Disruptive technology identification research based on one-class SVM: a case study in the field of gene editing

The identification of disruptive technologies is paramount for the expansion into novel markets and the creation of innovative value paradigms. Despite the importance of this task, current methodologies for identifying such technologies are predominantly qualitative, which are often hindered by a dearth of empirical data and a predisposition towards subjectivity. This paper proposes a paradigm shift by employing patent text data and utilizing Latent Dirichlet Allocation (LDA) for data preprocessing to extract and analyze technical themes from patent abstracts. The research innovatively posits the problem of identifying disruptive technologies as an anomaly detection issue and addresses it through the application of a one-class Support Vector Machine (SVM) algorithm. In the empirical component of the study, the field of gene editing is selected as a case study to validate the proposed method’s efficacy and practicality. The validation is conducted through a multifaceted analytical framework, which includes: an examination of the distribution and proportion of patent topics, a comprehensive market analysis, and a comparative analysis of the proposed model against existing methodologies. The results of this study not only substantiate the potential of the proposed approach in accurately detecting technologies with disruptive potential but also highlight its robustness across various analytical dimensions.

MOID: Many-to-One Patent Graph Embedding Base Infringement Detection Model

With the increasing number of patent applications over the years, instances of patent infringement cases have become more frequent. However, traditional manual patent infringement detection models are no longer suitable for large-scale infringement detection. Existing automated models mainly focus on detecting one-to-one patent infringements, but neglect the many-to-one scenarios. The many-to-one patent infringement detection model faces some major challenges. First, the diversity of patent domains, complexity of content and ambiguity of features make it difficult to extract and represent patent features. Second, patent infringement detection relies on the correlation between patents and the comparison of contextual information as the key factors, but modeling the process and drawing conclusions present challenges. To address these challenges, we propose a many-to-one patent graph (MPG) embedding base infringement detection model. Our model extracts the relationship between keywords and patents, as well as association relation between keywords from many-to-one patent texts (MPTs), to construct a MPG. We obtain patent infringement features through graph embedding of MPG. By using these embedding features as input, the many-to-one infringement detection (MOID) model outputs the conclusion on whether a patent is infringed or not. The comparative experimental results indicate that our model improves accuracy, precision and F-measure by 3.81%, 11.82% and 5.37%, respectively, when compared to the state-of-the-art method.

METHODS OF PATENT DOCUMENTATION ANALYSIS

Intellectual search of patent text is an important task that requires knowledge of the subject area. In today's world, when the growing pace of technological innovation is already leading to fundamental changes, with entering the fourth industrial revolution Industry (4.0), the development of artificial intelligence and the virtual economy, the analysis of patent documentation is becoming an increasingly important task, therefore, traditional approaches to text analysis are insufficient for comprehensive study of the patent text. This article discusses methods of patent document analysis, including topic identification, text segmentation, information display, feature selection, clustering, and summary selection, etc. The described methods can be used to better understand the relationships between different technologies and their priorities and to assess competition and opportunities for their use in the market. Also considered are the main challenges and problems associated with the analysis of patent documentation, such as the limited amount of data and the heterogeneity of their structure. The article's conclusions can be useful not only for the linguistic analysis of patent documentation, but also for researchers, engineers, managers and other specialists working in the field of innovation and technology development.

How to reach new technological specialisation: The role of related technological capabilities and complementary technological linkages

Technological specialisation is critical for regional development. One region can specialise in new technologies at a lower cost and faster speed by relating to specialised local technologies. Therefore, combined with the theory of path branching, we propose the research hypotheses that related technological capabilities and complementary technological linkages play important roles in the new technological specialisation. Moreover, we perform empirical analysis and test our hypotheses using patent text data on listed firms in China from 2000 to 2020. The results show that the impact of related technological capabilities on new technological specialisation is significantly positive. There is an optimal number of interregional technological linkages that promote new technological specialisation. Too many or too few are detrimental to the specialisation of new technologies. These results remain robust. Further investigation reveals that related technological capabilities and complementary technological linkages can work together to push for new technological specialisation throughout the sample period. However, this is weak in the short term or even insignificant in the central and western regions of China. These findings provide new insights and policy implications for emerging markets regarding reaching new technological specialisation by strengthening regional technological capabilities and collaboration.

Patent text mining based hydrogen energy technology evolution path identification

With the rise of the hydrogen energy industry, countries worldwide have introduced relevant policies to elevate the development of the hydrogen energy industry to the height of national energy strategy, and hydrogen energy will become a significant factor affecting the future industrial structure and national competitiveness. The path of technology evolution can trace the history of technology development and predict the trend of technology development, which can help to grasp the first opportunity of technological innovation and lead the direction of technological development. However, there are fewer studies on technology evolution in hydrogen energy. The existing studies mainly use bibliometrics to mine a single technology, and the studies using text mining suffer from the defect that keywords are difficult to decipher. Given this, this paper conducts research on identifying the technology evolution path in the field of hydrogen energy, adopts text mining methods to mine patent data, and introduces document vectorization and phrase mining algorithms to improve the mining depth while increasing the interpretability of the results. It is found that the hydrogen energy industry is in a period of rapid development; the succession and integration of hydrogen energy technologies are good, and the technological development trend focuses on hydrogen storage materials, hydrogen fuel cell vehicles, and green hydrogen preparation.

Automatic summarization of long text of Chinese patents based on PatBertsum model

Aiming at the problems of traditional text abstract extraction algorithms for processing long text of Chinese patents and unsatisfactory results of long abstract generation, the PatBertSum algorithm is proposed, which enables the algorithm to process long (more than 1500 words) patent text with high efficiency and generate high-quality long (more than 200 words) text summaries. The method is based on the improved BertSum algorithm model, using the new CLTPDS patented text dataset, processing long texts by Head-Tail, transforming Chinese input representations, generating sentence vectors using a pre-trained model, and capturing internal text features and text structure features to extract summaries. Experimentally, this paper demonstrates that the method has improved the recall and F-value of ROUGE by more than 8 percentage points compared with existing methods.

Demand Shocks, Procurement Policies, and the Nature of Medical Innovation: Evidence from Wartime Prosthetic Device Patents

Abstract We show that the demand shocks associated with the U.S. Civil War and World War I led to substantial increases in prosthetic device patenting (relative to patenting in other medical and mechanical technology classes). Through analyses of patent texts, we find that the Civil War led inventors to focus on production process improvements, while World War I did not. Further, we find that inventors emphasized dimensions of product quality that aligned with differences in buyers' preferences across wars. Alongside evidence from the historical record, these findings imply that procurement environments can significantly shape the scientific problems with which inventors engage.

Разработка программного модуля семантичского анализа патентного массива

While designing the software module, the authors developed an algorithm for parsing the texts of patents and patent applications, an algorithm for searching patents-analogues based on full-text search using Amazon Twinword technology, an algorithm for searching patents-analogues based on key phrases identified applying Amazon Comprehend technology, clustering the patent array. The software was developed to cluster the patent array and speed up the patent office examiner’s work by searching for key phrases and patent analogues

Measuring Patent Similarity Based on Text Mining and Image Recognition

Patent application is one of the important ways to protect innovation achievements that have great commercial value for enterprises; it is the initial step for enterprises to set the business development track, as well as a powerful means to protect their core competitiveness. The emergence of a large amount of patent data makes the effective detection of patent data difficult, and patent infringement cases occur frequently. Manual measurement in patent detection is slow, costly, and subjective, and can only play an auxiliary role in measuring the validity of patents. Protecting the inventive achievements of patent holders and realizing more accurate and effective patent detection were the issues explored by academics. There are five main methods to measure patent similarity: clustering-based method, vector space model (VSM)-based method, subject–action–object (SAO) structure-based method, deep learning-based method, and patent structure-based method. To solve this problem, this paper proposes a calculation method to fuse the similarity of patent text and image. Firstly, the SAO structure extraction technique is used for the patent text to obtain the effective content of the text, and the SAO structure is compared for similarity; secondly, the patent image information is extracted and compared; finally, the patent similarity is obtained by fusing the two aspects of information. The feasibility and effectiveness of the scheme are proven by studying a large number of patent similarity cases in the field of mechanical structures.

Acquisition Method of User Requirements for Complex Products Based on Data Mining

The vigorous development of big data technology has changed the traditional user requirement acquisition mode of the manufacturing industry. Based on data mining, manufacturing enterprises have the innovation ability to respond quickly to market changes and user requirements. However, in the stage of complex product innovation design, a large amount of design data has not been effectively used, and there are some problems of low efficiency and lack of objectivity of user survey. Therefore, this paper proposes an acquisition method of user requirements based on patent data mining. By constructing a patent data knowledge base, this method combines the Latent Dirichlet Allocation topic model and a K-means algorithm to cluster patent text data to realize the mining of key functional requirements of products. Then, the importance of demand is determined by rough set theory, and the rationality of demand is verified by user importance performance analysis. In this paper, the proposed method is explained and verified by mining the machine tool patent data in CNKI. The results show that this method can effectively improve the efficiency and accuracy of user requirements acquisition, expand the innovative design approach of existing machine tool products, and be applied to other complex product fields with strong versatility.

Introducing DynaPTI-constructing a dynamic patent technology indicator using text mining and machine learning.

Patent data is an established source of information for both scientific research and corporate intelligence. Yet, most patent-based technology indicators fail to consider firm-level dynamics regarding their technological quality and technological activity. Accordingly, these indicators are unlikely to deliver an unbiased view on the current state of firm-level innovation and are thus incomplete tools for researchers and corporate intelligence practitioners. In this paper, we develop DynaPTI, an indicator that tackles this particular shortcoming of existing patent-based measures. Our proposed framework extends the literature by incorporating a dynamic component and is built upon an index-based comparison of firms. Furthermore, we use machine-learning techniques to enrich our indicator with textual information from patent texts. Together, these features allow our proposed framework to provide precise and up-to-date assessments about firm-level innovation activities. To present an exemplary implementation of the framework, we provide an empirical application to companies from the wind energy sector and compare our results to alternatives. Our corresponding findings suggest that our approach can generate valuable insights that are complementary to existing approaches, particularly regarding the identification of recently emerging, innovation-overperformers in a particular technological field.

Identifying and Predicting Trends of Disruptive Technologies: An Empirical Study Based on Text Mining and Time Series Forecasting

Disruptive technologies are related to a country’s competitiveness and international status. Accurately identifying and predicting the trends in disruptive technologies through scientific methods can effectively grasp the dynamics of technological development, adjust the national science and technology strategic layout, and better seize the high ground in international competition. Based on patent text data, this paper uses the improved LDA2Vec model combined with relevant indicators to identify the main topics in disruptive technologies, and predicts and analyzes the development trend through the establishment of an ARIMA model. Taking the energy technology field as an example, the main topics and development trends concerning disruptive technologies in this field are obtained. The study found that ten technologies, including energy storage technology, energy internet management technology, and offshore wind energy technology, are disruptive technologies in the energy technology field, and the development speed of energy storage technology is the fastest. To verify the correctness of the conclusion, this paper compares the results with artificial verification methods such as expert interviews and document verification, and finds that the two are basically consistent, thus verifying the effectiveness and feasibility of the proposed method.

Mining semantic features in patent text for financial distress prediction

Financial distress prediction has been a popular topic over the decades. Most studies have used accounting features from financial statements to predict financial distress. Compared to listed companies, unlisted public companies have longer financial disclosure cycles, less required disclosure of market trading information, and higher financial risk. However, they can also have a strong ability to innovate and great growth potential, attributes that cannot be fully reflected in financial statements. In this study, as a supplement to accounting features, we propose a framework for mining the statistical features and semantic features in patent text by comprehensively analyzing the patent's structured information, abstract, claims, citations, and specifications. The results of empirical evaluation confirm that patent features contain incremental information related to financial distress. This research broadens the feature space of financial distress research and expands the research on patent text. It also provides decision support for banks approving loans, investment decision-making, and patent pledges.

Intellectual Property Information Intelligent System Based on Content Data Mining and ID4 Algorithm

This paper studies the intellectual property information system based on content data mining and ID4 algorithm. First, data mining technology is applied to patent information analysis, such as using clustering algorithm to mine patent text, and using association rules to mine patent inventors. The system can analyze the structure of DOCDB patent data files, extract relevant patent information, and store the processed data in the database. The experimental results show that the system can efficiently process patent data and effectively improve the automation level of patent preprocessing. Use support vector machines, naive Bayes, and radial basis neural networks to classify and test patent samples.

Technology opportunity discovery of proton exchange membrane fuel cell from the perspective of papers-patents correlation analysis

Papers and patents can respectively present the latest progress of scientific research and technological development. Combining the two for correlation analysis has certain reference significance for technology opportunity discovery. Therefore, this paper proposes a method for technology opportunity discovery of proton exchange membrane fuel cell from the perspective of papers-patents correlation analysis. Firstly, papers and patent data are collected from papers and patent databases and the LDA model is applied to extract paper and patent topics. Secondly, the indicators of topic heat and novelty are calculated to select high-value paper topics. After that, the similarity analysis between high-value paper topics and all patent topics is considered to select the patent topics with development potential. Finally, the patent texts with high relevance to these patent topics with development potential are positioned. The technical keywords in them are extracted by text mining tool and classified into innovation dimensions according to their attributes and the TEMPEST model. After that, the innovation rules in the SCAMPER model are introduced to combine with the technical keywords in each innovation dimension to identify specific technology opportunities. An empirical study on the technology of proton exchange membrane fuel cell provides sufficient evidence of the method’s ability to discover technology opportunities. This paper contributes by generating four specific technology opportunities, which can make up for the existing defects of proton exchange membrane fuel cells and provide useful guidance for companies to carry out technological innovation.

Copyright Protection of Computer Software Deep Learning based Patent Text Clustering

Computer-Aided Design and Applications is an international journal on the applications of CAD and CAM. It publishes papers in the general domain of CAD plus in emerging fields like bio-CAD, nano-CAD, soft-CAD, garment-CAD, PLM, PDM, CAD data mining, CAD and the internet, CAD education, genetic algorithms and CAD engines. The journal is aimed at all developers and users of CAD technology to ptovide CAD solutions for various stages of design and manufacturing. The journal publishes all about Computer-Aided Design and Computer-Aided technologies.

The Evolution of Artificial Intelligence in the Digital Economy: An Application of the Potential Dirichlet Allocation Model

The most critical driver of the digital economy comes from breakthroughs in cutting-edge technologies such as artificial intelligence. In order to promote technological innovation and layout in the field of artificial intelligence, this paper analyzes the patent text of artificial intelligence technology using the LDA topic model from the perspective of the patent technology subject based on Derwent patent data. The results reveal that AI technology is upgraded from chips, sensing, and algorithms to innovative platforms and intelligent applications. Proposed countermeasures are necessary to advance the digitalization of the global economy and to achieve economic globalization in terms of industrial integration, building ecological systems, and strengthening independent innovation.

Patent Text Classification based on Deep Learning and Vocabulary Network

Patent documents are a special long text format, and traditional deep learning methods have insufficient feature extraction ability, which results in a weaker classification effect than ordinary text. Based on this, this paper constructs a text feature extraction method based on the lexical network, according to the inner relation between words and classification. Firstly, the inner relationship between words and classification was obtained from linear and probability dimensions and the lexical network were constructed. Secondly, the lexical network is fused with the features extracted from the deep learning model. Finally, the fusion features are trained in the original model to get the final classification result. T This method is a classification enhancement method that can classify patent text alone or enhance the accuracy of various types of neural networks in patent text classification. Experimental results demonstrate that the accuracy of BERT combined with lexical network method is as high as 82.73%, and the accuracy of lexical network method combined with CNN and LSTM is increased by 2.19% and 2.25% respectively. In addition, it was demonstrated that the lexical network feature extraction method accelerated the convergence speed of the model during training and improved the classification ability of the model in Chinese patent texts.