World Patent Information最新文献

{"title":"Assessing the standard-essentiality of 5G technology patents by means of generative artificial intelligence","authors":"Andre Herzberg","doi":"10.1016/j.wpi.2025.102363","DOIUrl":"10.1016/j.wpi.2025.102363","url":null,"abstract":"<div><div>In telecommunication technology, identifying standard-essential patents (SEPs) plays a crucial role in the management of intellectual property. This technology is regulated by technical standards that are largely based on the content of SEPs. These patents are declared standard-essential by their owners because they contain elements of a technical standard. The declaration process leaves room for over- and under-declaration, which entails risks for organizations. This paper focuses on the question of how generative artificial intelligence can be used to assess the standard-essentiality of 5G technology patents. For this purpose, the standard-essentiality is assessed using different prompts with four Large Language Models (LLMs) in two variants. In the first variant, the LLM results are generated by a rather simple prompt and compared with an approach based on unsupervised and supervised machine learning. The result shows that large LLMs are capable of assessing the standard-essentiality. In the second variant, the best-performing LLM is selected and the prompt is expanded to include selected parts of a technical standard. While the assessment results remain largely the same, the LLM is now able to explain in which detail a patent is part of a standard. This has several implications for patent evaluation, licensing and litigation strategies.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102363"},"PeriodicalIF":2.2,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143922957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improved multi-label hierarchical patent classification using LLMs","authors":"Bardia Rafieian,&nbsp;Pere-Pau Vázquez","doi":"10.1016/j.wpi.2025.102356","DOIUrl":"10.1016/j.wpi.2025.102356","url":null,"abstract":"<div><div>Classifying multi-label documents has always been a challenging task, especially when the labels follow a hierarchical structure. This complexity increases the difficulty of accurately predicting multiple interrelated labels, often limiting the overall classification performance. To address these challenges and improve the accuracy of hierarchical multi-label classification systems, we introduce a novel pipeline leveraging large language models (LLMs). By incorporating quantization techniques and optimizing weight updates through smaller matrices, we improve computational efficiency and scalability. Our approach demonstrates an improvement in accuracy compared to current state-of-the-art models. In this work, we apply this pipeline to patent documents, focusing on the multi-label hierarchical text classification problem using a transformer-based architecture. The hierarchical structure of the CPC (Cooperative Patent Classification) labels is preserved through a graph-based taxonomy, which enables more effective processing of patent categories. Our model is trained and evaluated on the USPTO-70k dataset, and we achieve substantial improvements across various metrics, including precision, recall, F1-score, and AUC.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102356"},"PeriodicalIF":2.2,"publicationDate":"2025-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143927404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"How the magic happens: Patenting work at a multinational manufacturing company","authors":"Mimmi Hanson","doi":"10.1016/j.wpi.2025.102362","DOIUrl":"10.1016/j.wpi.2025.102362","url":null,"abstract":"<div><div>Patents have become increasingly important in the global economy in recent decades. Today, patents are used for various purposes, far beyond the initial role of preventing imitation of new inventions introduced into society. In the work to develop solutions for more sustainable societies, patents have the potential to facilitate the necessary collaboration and sharing of technology. To complement the many studies on using patents for these purposes, this paper examines the intra-organisational practices with which patent strategies are implemented. The concept of boundary objects is introduced to analyse patenting work and its coordination across different sites. The paper makes three contributions to the literature. First, strategy work is a necessary but not a sufficient part of patenting work; implementing a strategy relies on all patenting work in the organisation. Second, In the organisation of patenting work, the focus should be on identifying relevant communities of experts and not simply organisational units. Third, \"patents\" as boundary objects can coordinate discourse around patents, which creates motivation and commitment to engage in patenting work. It is, however, not strong enough to coordinate the work itself; this is instead coordinated by patent-related artefacts that travel between the nodes.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102362"},"PeriodicalIF":2.2,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143881677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Encoder models at the European Patent Office: Pre-training and use cases","authors":"Volker D. Hähnke,&nbsp;Arnaud Wéry,&nbsp;Matthias Wirth,&nbsp;Alexander Klenner-Bajaja","doi":"10.1016/j.wpi.2025.102360","DOIUrl":"10.1016/j.wpi.2025.102360","url":null,"abstract":"<div><div>Patents are organized using systems of technical concepts like the Cooperative Patent Classification. Classification information is extremely valuable for patent professionals, particularly for patent search. Language models have proven useful in Natural Language Processing tasks, including document classification. Generally, pre-training on a domain is essential for optimal downstream performance. Currently, there are no models pre-trained on patents with sequence length above 512. We pre-trained a RoBERTa model with sequence length 1024, increasing the fully covered claims sections from 12% to 53%. It has a ‘base’ configuration, reducing free parameters compared to ‘large’ models in the patent domain three-fold. We fine-tuned the model on classification tasks in the CPC, up to leaf level. Our tokenizer produces sequences on average 5% and up to 10% shorter than the general English RoBERTa tokenizer. With our pre-trained ‘base’ size model, we reach classification performance better than general English models, comparable to ‘large’ models pre-trained on patents. On the finest CPC granularity, 88% of test documents have at least one ground truth symbol in the top 10 predictions. Our CPC prediction models and data sets are publicly accessible. With the described procedures, we can periodically repeat pre-training and fine-tuning to cope with drift effects.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102360"},"PeriodicalIF":2.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamics of regional intellectual property systems in China: A spatiotemporal synergy analysis","authors":"Li Chen ,&nbsp;Liang Gao ,&nbsp;Sibei Sheng","doi":"10.1016/j.wpi.2025.102359","DOIUrl":"10.1016/j.wpi.2025.102359","url":null,"abstract":"<div><div>In the context of a burgeoning scientific and technological revolution and evolving norms, the push towards innovation-driven development has become crucial for achieving high-quality national growth. Such development is essential for enhancing the coordinated evolution of regional intellectual property (IP) systems. This study analyses data from 30 provincial units in China to develop an evaluation index system for regional IP synergy, encompassing the following four subsystems: IP creation, utilization, protection, and service. Using the entropy method, this study assesses the organizational level within each subsystem from 2009 to 2022 and evaluates the degree of coordinated development of regional IP systems. Furthermore, this work examines interregional disparities through the coefficient of variation, the Gini coefficient, and the Theil index. The findings reveal that the synergy of China's regional IP systems is characterized by growth fluctuations and significant regional disparities. Low levels of synergy can impede the enhancement of regional IP capabilities and reduce the efficiency of IP output. Disparities in synergy levels among regions are the main obstacles to connectivity and cooperation within the relevant network.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102359"},"PeriodicalIF":2.2,"publicationDate":"2025-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143850147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Patenting telemedicine tools: A cross-country analysis of technologies related to remote patient monitoring","authors":"Gazala Parveen,&nbsp;Padmavati Manchikanti","doi":"10.1016/j.wpi.2025.102361","DOIUrl":"10.1016/j.wpi.2025.102361","url":null,"abstract":"<div><div>Telemedicine has been practised since digital technology emerged in the mid-to-late 20th century. It evolved with technological advancements such as satellite communication in the 1960s, the internet in the 1990s and mobile health applications in the 2000s. Today, telemedicine forms a sub-set of digital health. In telemedicine, healthcare professionals provide medical services through information and communication technologies. The patenting of telemedicine tools is quite active, encompassing advancements in software applications, medical devices, and integrated systems for remote diagnosis, monitoring, and treatment. Effective protection of intellectual property for telemedicine tools relies on organised patent management and precise claim drafting. A study on the filing trends of the patents related to telemedicine tools and patent prosecution will give a better understanding of issues related to the patenting of such technologies. It highlights new developments that are expanding the scope of patent claims, particularly the increasing integration of telemedicine and software-enabled medical devices.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102361"},"PeriodicalIF":2.2,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143826182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Expanding the concept of drug lifecycle management to chimeric antigen receptor T-cell products through product-patent linkage analysis","authors":"Yasuaki Kawai,&nbsp;Shingo Kano","doi":"10.1016/j.wpi.2025.102357","DOIUrl":"10.1016/j.wpi.2025.102357","url":null,"abstract":"<div><div>Chimeric antigen receptor T (CAR-T) cell therapies have been actively developed, and five CAR-T products have been commercialized in Japan. Due to the ongoing development of CAR-T cell therapies, including next-generation variants, the patent landscape is expected to become increasingly complex. Therefore, understanding patent strategies for each CAR-T product is essential.</div><div>In the pharmaceutical industry, lifecycle management (LCM) centered on regulatory and patent protection has been implemented to maximize product value. While studies have reported CAR-T patents through patent landscape analysis to gain insights into the overall CAR-T technology, there is a lack of research on product-related patents for CAR-T products. As a result, the foundational knowledge regarding the LCM of CAR-T products remains unclear.</div><div>Therefore, we identified product-patent linkages for CAR-T products in the Japanese market by combining patent term extension (PTE) data with publicly available data and assessed the applicability of drug LCM to CAR-T products. Our identification of precise product-patent linkages revealed that all CAR-T products met the criteria for drug LCM. This study suggests that LCM activities can be implemented for CAR-T products and that the concept of drug LCM can be expanded to CAR-T products.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":""},"PeriodicalIF":2.2,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143799537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Do large language models understand patents? Enhancing patent classification through AI-generated summaries","authors":"Naoya Yoshikawa ,&nbsp;Ralf Krestel","doi":"10.1016/j.wpi.2025.102353","DOIUrl":"10.1016/j.wpi.2025.102353","url":null,"abstract":"<div><div>Patent classification plays a crucial role in intellectual property management, but remains a challenging task due to the complexity of patent documents. This study explores a novel approach to enhance automatic patent classification by leveraging summaries generated by large language models (LLMs). Our approach involves using the GPT-3.5-turbo model to create concise summaries from different sections of patent texts, which are then used to fine-tune the RoBERTa and XLM-RoBERTa models for classification tasks. We conducted experiments on English and Japanese patent documents using two datasets: the well-established USPTO-70k and the newly developed JPO-70k, that we specifically created for this study.</div><div>Our findings show that models trained on AI-generated summaries – particularly those derived from patent claims or detailed descriptions – outperform models trained on original abstracts in both subclass-level multi-label classification and subgroup-level single-label classification. In particular, using detailed description summaries improved the micro-average F1 score for subclass-level classification by 2.9 points on the USPTO-70k and 3.0 points on the JPO-70k, compared to using original abstracts.</div><div>These results indicate that LLM-generated summaries effectively capture information relevant to patent classification from various sections of patent texts, offering a promising approach to enhance the accuracy and efficiency of patent classification across different languages.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102353"},"PeriodicalIF":2.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143776335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrating Generative Artificial Intelligence techniques into technology function matrix analysis","authors":"Huei-Yu Wang ,&nbsp;Shu-Hao Chang ,&nbsp;Chia-Yi Chuang","doi":"10.1016/j.wpi.2025.102352","DOIUrl":"10.1016/j.wpi.2025.102352","url":null,"abstract":"<div><div>This study proposes a novel method for automating the construction of technology-function matrices using generative artificial intelligence (GAI), specifically focusing on quantum technologies. By leveraging GAI to analyze International Patent Classification (IPC) definitions and benchmark reports, we developed a system that rapidly generates technology-function matrices, significantly reducing the time required for manual analysis. The method was applied to 2,399 quantum technology patents from 2023 to March 2024, covering four key areas: secure communications, computing, quantum simulators, and sensors. This approach not only aids government agencies in identifying new technological opportunities but also facilitates the industrialization of potential technologies. By combining GAI with established analytical frameworks, this study contributes to both the theoretical understanding and practical application of patent analysis in emerging fields.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102352"},"PeriodicalIF":2.2,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143759113","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Knowledge flows in technology-intensive publicly listed company - Evidence from Chinese patent citation data","authors":"Shi Chen ,&nbsp;Yifa Wang","doi":"10.1016/j.wpi.2025.102354","DOIUrl":"10.1016/j.wpi.2025.102354","url":null,"abstract":"<div><div>This empirical study utilizes patent citations from technology-intensive publicly listed companies in China between 2000 and 2019 to analyze the current state of knowledge flow within and among these companies. While capital markets are crucial for innovation, the extent to which publicly listed firms facilitate knowledge flow remains unclear. This research delineates the circulation of technological knowledge both intra- and inter-company, across similar and disparate industries, and between listed companies and non-listed innovative entities. The findings indicate a pronounced tendency among technology-intensive listed companies to cite their patents extensively. Self-citations comprise nearly 60 % of total citations, suggesting that technological knowledge primarily circulates within individual companies. Furthermore, the exchange of technological knowledge among different listed companies within the same industry is notably sparse, with only a fractional increase in the frequency of knowledge flows within the industry compared to across industry boundaries. Predominantly, the technological knowledge that technology-intensive listed companies acquire from non-listed innovative entities stems from domestic unlisted companies, with foreign entities and universities contributing to a lesser extent. When examining the spillover of technological knowledge to non-listed innovative entities, it is observed that other non-listed companies predominantly absorb such knowledge, with universities and individual innovators receiving lesser proportions. Finally, this study is significant as it provides empirical evidence on the flow of technological knowledge within and between publicly listed technology-intensive companies in China, revealing the dominance of self-citations and limited cross-company knowledge exchange. By analyzing patent citation data, this research provides valuable insights into the interactions between listed companies and non-listed innovative entities. The findings highlight the significant role of non-listed firms, universities, and foreign entities in shaping technological development. Strengthening these connections can further foster innovation and en hance knowledge diffusion across sectors.</div></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"81 ","pages":"Article 102354"},"PeriodicalIF":2.2,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143734634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}