An Approval Time Prediction Method Based on Patent Characteristics
Xiang Shuxuan1, Li Rui2
1.Laboratory of Data Intelligence and Interdisciplinary Innovation, Nanjing University, Nanjing 210023 2.Institute for Disaster Management and Reconstruction, Sichuan University, Chengdu 610207
摘要创新主体预测竞争对手或自身专利能否获批及获批速度，是一项重要的竞争情报工作。审批耗时会占用专利的有效期，有的发明专利审批耗时达8年甚至12年。审批耗时越长，对从申请日起算最长20年的专利有效期的折损就越大，对专利价值的影响也越大。本文旨在构建一种情报学方法，通过挖掘专利文献的系列特征形成预测模型，用于预测专利能否获批以及获批的速度。整个研究包括两个部分逻辑内容，即获批预测和速度预测。首先，应用相关分析与Cox比例风险回归模型对所选特征进行检验。其次，在此基础上，针对前面提取到的技术内容量属性、技术结构属性、技术功能属性、技术概念明确度属性以及申请人发明人属性等系列特征，使用辅助学习的方法，利用审查结果与审查周期的关联信息构建专利获批速度预测模型（patent approval time prediction model，MACP）。研究结果显示，基于辅助任务组合的MACP较已有的基线模型表现更佳。由于MACP模型有效地学习与利用了更多的专利审查过程知识，降低了对数据量的依赖，能取得更好的预测效果。
向姝璇, 李睿. 基于专利文献特征的专利获批预测模型[J]. 情报学报, 2024, 43(12): 1467-1482.
Xiang Shuxuan, Li Rui. An Approval Time Prediction Method Based on Patent Characteristics. 情报学报, 2024, 43(12): 1467-1482.
1 Bagna E, Cotta Ramusino E, Denicolai S. Innovation through patents and intangible assets: effects on growth and profitability of European companies[J]. Journal of Open Innovation: Technology, Market, and Complexity, 2021, 7(4): 220. 2 Mitra-Kahn B, Marco A, Carley M, et al. Patent backlogs, inventories and pendency: an international framework[R/OL]. IPO- & USPTO Joint Report. Newport: The Intellectual Property Office, (2013-06-26). https://assets.publishing.service.gov.uk/media/5a7e 51f040f0b6230268a97f/ipresearch-uspatlog-201306.pdf. 3 Burke P F, Reitzig M. Measuring patent assessment quality—analyzing the degree and kind of (in) consistency in patent offices’ decision making[J]. Research Policy, 2007, 36(9): 1404-1430. 4 Harhoff D, Wagner S. The duration of patent examination at the European patent office[J]. Management Science, 2009, 55(12): 1969-1984. 5 Frakes M D, Wasserman M F. Is the time allocated to review patent applications inducing examiners to grant invalid patents? Evidence from microlevel application data[J]. Review of Economics and Statistics, 2017, 99(3): 550-563. 6 Arts S, Hou J N, Gomez J C. Natural language processing to identify the creation and impact of new technologies in patent text: code, data, and new measures[J]. Research Policy, 2021, 50(2): 104144. 7 Hido S, Suzuki S, Nishiyama R, et al. Modeling patent quality: a system for large-scale patentability analysis using text mining[J]. Journal of Information Processing, 2012, 20(3): 655-666. 8 Sun Z, Wright B D. Citations backward and forward: insights into the patent examiner's role[J]. Research Policy, 2022, 51(7): 104517. 9 Zhang G P, Xiong L B, Duan H B, et al. Obtaining certainty vs. creating uncertainty: does firms' patent filing strategy work as expected?[J]. Technological Forecasting and Social Change, 2020, 160: 120234. 10 Jell F, Henkel J, Hoisl K. The relationship between patenting motives and pendency durations[J]. SSRN Electronic Journal, 2013. DOI: 10.2139/ssrn.2406536. 11 欧桂燕, 庞娜, 吴江. 专利审查周期影响因素研究——以中国人工智能领域为例[J]. 数据分析与知识发现, 2022, 6(8): 20-30. 12 de Rassenfosse G, Dernis H, Guellec D, et al. The worldwide count of priority patents: a new indicator of inventive activity[J]. Research Policy, 2013, 42(3): 720-737. 13 Liegsalz J, Wagner S. Patent examination at the state intellectual property office in China[J]. Research Policy, 2013, 42(2): 552-563. 14 Yang D L. Pendency and grant ratios of invention patents: a comparative study of the US and China[J]. Research Policy, 2008, 37(6-7): 1035-1046. 15 Zhu K J, Malhotra S, Li Y H. Technological diversity of patent applications and decision pendency[J]. Research Policy, 2022, 51(1): 104364. 16 Hoock C, Brown A. Early certainty in patent cases involving by opposition proceedings[J]. World Patent Information, 2020, 61: 101948. 17 Langinier C, Marcoul P. Monetary and implicit incentives of patent examiners[J]. Journal of Economics and Business, 2020, 110: 105906. 18 Marco A C, Sarnoff J D, de Grazia C A W. Patent claims and patent scope[J]. Research Policy, 2019, 48(9): 103790. 19 Popp D, Juhl T, Johnson D K N. Time in purgatory: determinants of the grant lag for U.S. patent applications[J]. The B.E. Journal of Economic Analysis & Policy, 2004, 4(1): No.9518. 20 Golodne D M, da Silva A S. Managing patent examiners’ workload by using patent application volume data[J]. World Patent Information, 2022, 68: 102095. 21 Chi Y C, Wang H C. Establish a patent risk prediction model for emerging technologies using deep learning and data augmentation[J]. Advanced Engineering Informatics, 2022, 52: 101509. 22 Lai C Y, Hwang S Y, Wei C P. On the patent claim eligibility prediction using text mining techniques[C]// Proceedings of the 51st Annual Hawaii International Conference on System Sciences. Piscataway: IEEE, 2018: 587-596. 23 Jiang H X, Fan S K, Zhang N, et al. Deep learning for predicting patent application outcome: the fusion of text and network embeddings[J]. Journal of Informetrics, 2023, 17(2): 101402. 24 Schmitt V J, Walter L, Schnittker F C. Assessment of patentability by means of semantic patent analysis–a mathematical-logical approach[J]. World Patent Information, 2023, 73: 102182. 25 USPTO. First office action estimator[EB/OL]. [2024-07-10]. https://www.uspto.gov/learning-and-resources/statistics/first-office- action-estimator. 26 Dutt R, Krishna V. Forecasting the grant duration of a patent using predictive analytics[J]. International Journal of Computer Applications, 2019, 178(51): 1-7. 27 Dutt R, Rathi P, Krishna V. Novel mixed-encoding for forecasting patent grant duration[J]. World Patent Information, 2021, 64: 102007. 28 Zhang G J, Yu G, Feng Y Q, et al. Improving the publication delay model to characterize the patent granting process[J]. Scientometrics, 2017, 111(2): 621-637. 29 Mesmer H A, Cunningham J W, Hiebert E H. Toward a theoretical model of text complexity for the early grades: learning from the past, anticipating the future[J]. Reading Research Quarterly, 2012, 47(3): 235-258. 30 Li M, Kirby J R. The effects of vocabulary breadth and depth on English reading[J]. Applied Linguistics, 2015, 36(5): 611-634. 31 Zheng L Q, Huang R H, Hwang G J, et al. Measuring knowledge elaboration based on a computer-assisted knowledge map analytical approach to collaborative learning[J]. Journal of Educational Technology & Society, 2015, 18: 321-336. 32 Tong T W, Zhang K, He Z L, et al. What determines the duration of patent examination in China? An outcome-specific duration analysis of invention patent applications at SIPO[J]. Research Policy, 2018, 47(3): 583-591. 33 Ashtor J H. Modeling patent clarity[J]. Research Policy, 2022, 51(2): 104415. 34 朱桂龙, 王萧萧. 专利质量影响因素分析——基于专利引文结构新视角[J]. 中国科技论坛, 2019(9): 67-75. 35 Mortensen P S. Patentometrics as performance indicators for allocating research funding to universities[R/OL]. Aarhus: University of Aarhus, (2011-06-15) [2024-07-10]. https://pure.au.dk/ws/files/39714215/WP2011_1_Patentometrics_as_performance_indicators.pdf. 36 Novelli E. An examination of the antecedents and implications of patent scope[J]. Research Policy, 2015, 44(2): 493-507. 37 Marku E. Measuring innovation quality: a patent analysis[J]. IOSR Journal of Business and Management, 2018, 20(8): 51-58. 38 Ferrucci E, Lissoni F. Foreign inventors in Europe and the United States: diversity and patent quality[J]. Research Policy, 2019, 48(9): 103774. 39 van Zeebroeck N. The puzzle of patent value indicators[J]. Economics of Innovation and New Technology, 2011, 20(1): 33-62. 40 Wang Z Y, Xia L S, Yuan H P, et al. Principles, research status, and prospects of feature engineering for data-driven building energy prediction: a comprehensive review[J]. Journal of Building Engineering, 2022, 58: 105028. 41 Chandrashekar G, Sahin F. A survey on feature selection methods[J]. Computers & Electrical Engineering, 2014, 40(1): 16-28. 42 Cammarano A, Michelino F, Lamberti E, et al. Accumulated stock of knowledge and current search practices: the impact on patent quality[J]. Technological Forecasting and Social Change, 2017, 120: 204-222. 43 Du Y S, Czarnecki W M, Jayakumar S M, et al. Adapting auxiliary losses using gradient similarity[EB/OL]. (2020-11-25). https://arxiv.org/pdf/1812.02224. 44 Caruana R. Multitask learning[M]// Learning to Learn. Boston: Springer, 1998: 95-133. 45 Liu C Y, Wu H L, Lee C Y. The relationship between patent attributes and patent litigation: considering the moderating effects of managerial characteristics[J]. Asia Pacific Management Review, 2018, 23(2): 121-129.
