Research on Automated Generation and Evaluation of Patent ClaimsBased on GPT-4

Authors

• Junhua LI, Shanghai Maritime University Library(Archive), Shanghai 201306
• Qian YUAN, Shanghai Maritime University Library(Archive) , Shanghai 201306
• Xiang YAN, Shanghai Maritime University Library(Archive) , Shanghai 201306
• Changhong LV, Shanghai Maritime University Library(Archive) , Shanghai 201306

Keywords

patent claims; machine reading comprehension; GPT-4; Prompt; patent

Abstract

[Purpose/significance]This study aims to automatically generate claims using the GPT-4 model, in order to reduce the writing difficulty for inventor and improve the work efficiency and quality. [Method/process]The article constructs Prompts suitable for automatically generating patent claims and implements four prompting strategies: ZeroShot, Exact-Drafting, Stepwise-Claim, and Exact-Step Claim. By inputting patent specifications and technical disclosure documents into the GPT-4 model and using Prompts to guide its output, the automated generation of patent claims is achieved. The ROUGE and BERTScore evaluation metrics were used to assess the quality of the text, and the generated text was analyzed in comparison with the reference text from multiple dimensions, including the number of claims, text length, high-frequency words, keywords, and common collocations. Finally, the quality of the generated claim documents was evaluated through expert assessment in five aspects: clarity, consistency, relevance, professionalism, and completeness. [Result/ conclusion]Empirical research shows that the Exact-Step Claim prompting strategy significantly improves the quality of generated claims; moreover, claims generated based on patent specifications are more closely matched in the number of claims and text length with the reference texts, indicating that the application effect of the GPT-4 model in the field of natural language understanding and generation is closely related to the quality of the input text. This study provides an efficient and intelligent assistance method that contributes to the development of the patent text writing and review field. However, there are challenges, and further improvements are needed for the model to precisely understand complex technical terms and comply with patent regulations, as well as to explore how to optimize the model's ability to judge the number of claims and the length of the text.

First Page

95

Last Page

108

Submission Date

16-Apr-2024

Revision Date

17-Jun-2024

Acceptance Date

19-Jul-2024

Published Date

01-Jan-2025

Reference

[1] 中国人大网. 中华人民共和国专利法[EB/OL]. (2020-11-09) [2024-05-12]. http://www.npc.gov.cn/npc/c2/c30834/202011/t20201119_308800.html.
[2] 陈姝. 权利要求保护范围确定中重要问题的探析[J]. 中国发明与专利, 2020, 17 (06): 116-118.
[3] CHURNET D. Patent Claims Revisited[J]. Northwestern Journal of Technology and Intellectual Property, 2013, 11 (06): 501-520.
[4] WAGNER R P. Understanding Patent-Quality Mechanisms[J]. University of Pennsylvania Law Review, 2009, 157 (06): 2135-2173.
[5] 赵恩博. 专利权利要求书生成关键技术研究[D]. 哈尔滨: 哈尔滨工业大学, 2020.
[6] 中华人民共和国中央人民政府. 政府工作报告[EB/OL]. (2024-03-12) [2024-05-12]. https://www.gov.cn/gongbao/2024/issue_11246/202403/content_6941846.html.
[7] 南京理工大学. 全国首个知识产权大模型 (IP-GPT) 在南京发布[EB/OL]. (2023-09-10) [2024-05-12]. https://zscq.njust.edu.cn/ec/fd/c2201a322813/page.htm.
[8] 大连理工大学. “太保”: 为知识产权保驾护航[EB/OL]. (2023-12-16) [2024-05-12]. https://mp.weixin.qq.com/s/zmPpAWnxxIC0g1trD3sCrQ.
[9] 王为磊. 专利大模型的实践与知识问答探索[EB/OL]. (2024-05-06) [2024-05-12]. https://www.51cto.com/article/787639.html.
[10] 中国科学院知识产权信息. 韩国知识产权局将建立全球首个AI专利审查系统[EB/OL]. (2023-07-20) [2024-04-01]. https://mp.weixin.qq.com/s/peYtYFrKhPicedk0fKwrZA.
[11] 法律资讯. 美国专利商标局发布关于人工智能辅助发明的发明人指南[EB/OL]. (2024-02-12) [2024-04-01]. https://zhuanlan.zhihu.com/p/684587390.
[12] 中国科学院知识产权信息. 日本专利局新设人工智能顾问支持专利审查[EB/OL]. (2024-04-08) [2024-04-01]. https://mp.weixin.qq.com/s/IBVEexPogmSu84IC8ou13w.
[13] SCHANK R C, ABELSON R P. Scripts, Plans, Goals, and Understanding: An Inquiry into Human Knowledge Structures[M]. New York: Psychology Press, 1977.
[14] BERANT J, CHOU A, FROSTIG R, et al. Semantic Parsing on Freebase from Question-answer Pairs[C]//Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, 2013: 1533-1544.
[15] HERMANN K M, KOCISKY T, GREFENSTETTE E, et al. Teaching machines to read and comprehend [C]//Proceedings of the 28th International Conference on Neural Information Processing Systems, 2015: 1693-1701.
[16] DEVLIN J, CHANG M-W, LEE K, et al. Bert: Pre-training of Deep Bidirectional Transformers for Language Understanding[C]//Proceedings of NAACL-HLT, 2019: 4171-4186.
[17] PETERS M, NEUMANN M, IYYER M, et al. Deep Contextualized Word Representations[C]//Proceedings of NAACL-
HLT, 2018: 2227-2237.
[18] RADFORD A, NARASIMHAN K, SALIMANS T, et al. Improving Language Understanding by Generative pre-training[J/OL]. [2024-01-10]. https://www.mikecaptain.com/resources/pdf/GPT-1.pdf.
[19] 白如江, 陈启明, 张玉洁, 等. 基于ChatGPT+Prompt的专利技术功效实体自动生成研究[J]. 数据分析与知识发现, 2024, 8 (04): 14-25.
[20] 王奎芳, 吕璐成, 孙文君, 等. 基于大模型知识蒸馏的专利技术功效词自动抽取方法研究: 以车联网V2X领域为例[J/OL]. 数据分析与知识发现, 1-20[2024-09-05]. http://kns.cnki.net/kcms/detail/10.1478.G2.20240305.1406.004.html.
[21] ZHONG Q, DING L, LIU J, et al. Can ChatGPT Understand Too?A Comparative Study on ChatGPT and Fine-tuned BERT[DB/OL]. (2023-02-19) [2024-03-20] https://arxiv.org/abs/2302.10198.
[22] WEI X, CUI X, CHENG N, et al. Zero-Shot InformationExtraction ViaChatting with Chatgpt[DB/OL]. (2023-02-20) [2024-03-20]. https://arxiv.org/abs/2302.10205.
[23] JIAO W, WANG W, HUANG J-T, et al. Is ChatGPT A GoodTranslator?A Preliminary Study[DB/OL]. (2023-01-20) [2024-03-25]. https://arxiv.org/abs/2301.08745.
[24] WANG J, LIANG Y, MENG F, et al. Cross-Lingual Summarization Via Chatgpt[DB/OL]. [2024-03-25]. https://arxiv.org/pdf/2302.14229v1.
[25] PETRUZZI J D, MASON R M. Machine For Drafting A Patent Application and Process for Doing Same: US6049811[P/OL]. 2000-04-11[2024-05-12]. https://ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6049811.
[26] SHEREMETYEVA S, NIRENBURG S, NIRENBURG I. Generating Patent Claims from Interactive Input [C]//Proceedings of the Eighth International Natural Language Generation Workshop, 1996: 61-70.
[27] KNIGHT K, SCHICR I C, PRIYADARSHI J. Machine Learning Model for Computer-Generated Patent Applications to Provide Support for Individual Claim Features in a Specification, US10713443[P/OL]. 2020-07-14[2024-05-12]. https://ppubs.uspto.gov/dirsearch-public/print/downloadPdf/10713443.
[28] LEE J-S. Personalized Patent Claim Generation and Measurement[DB/OL]. (2019-12-07) [2024-03-20]. https://arxiv.org/abs/1912.03502v2.
[29] LEE J-S, HSIANG J. Patent Claim Generation by Fine-Tuning OpenAI GPT-2[J]. World Patent Information, 2020 (62): 101983.
[30] MORENO S. Transformers-Based Abstractive Summarization for the Generation of Patent Claims[D]. Torino: Politecnico di Torino, 2023.
[31] LIU P, YUAN W, FU J, et al. Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing[J]. ACM Comput Surv, 2023, 55 (09): 1-35.
[32] LIN C-Y. Rouge: A Package for Automatic Evaluation of Summaries[C]//Proceedings of the Text Summarization Branches Out, 2004: 1-8.
[33] ZHANG T, KISHORE V, WU F, et al. Bertscore: Evaluating Text Generation with Bert[DB/OL]. (2019-04-21) [2024-03-20]. https://arxiv.org/abs/1904.09675.
[34] 刘竹辰, 陈浩, 于艳华, 等. 词位置分布加权TextRank的关键词提取[J]. 数据分析与知识发现, 2018, 2 (09): 74-79.
[35] 王巍洁, 任慧玲, 李晓瑛, 等. 融合汉字多语义与文本统计特征的中文医学新词发现研究[J]. 图书情报工作, 2024, 68 (06): 119-128.

Digital Object Identifier (DOI)

10.19809/j.cnki.kjqbyj.2025.01.009

Recommended Citation

LI, Junhua; YUAN, Qian; YAN, Xiang; and LV, Changhong (2025) "Research on Automated Generation and Evaluation of Patent ClaimsBased on GPT-4," Journal of Scientific Information Research: Vol. 7: Iss. 1, Article 9.
DOI: 10.19809/j.cnki.kjqbyj.2025.01.009
Available at: https://eng.kjqbyj.com/journal/vol7/iss1/9