[目的/意义] 从专利分类修订角度研究技术演化，为技术演化研究提供新思路。[方法/过程] 首先根据IPC分类表H部2009-2018年的修订情况，总结出新增分类、删除分类、类内转移分类、类间转移分类4种修订类型。其次针对分类修订后产生的过档文献提出基于Word2vec+TextCNN模型的过档文献再分类方法，使新旧版分类表通过再分类专利产生衔接。最后结合H部2009年-2018年被修订分类及再分类专利进行技术演化初步探索。[结果/结论] 专利再分类模型可有效解决过档文献问题，为专利再分类工作提供参考，同时可衔接新旧版专利分类表；结合IPC分类修订及再分类专利可分析分类修订中的主要技术演化方向，为技术演化研究提供新视角。
[Purpose/significance] This article attempts to study technological evolution from the perspective of patent classification revisions, and provides new ideas for technological evolution research. [Method/process] First, according to the 2009-2018 revisions of the H part of the IPC classification table, four types of revisions were summarized: new classification, deletion classification, intra-class transfer classification, and inter-class transfer classification. Secondly, for the archived documents generated after classification revision, a reclassification method of patent archived documents based on the Word2vec+TextCNN model was proposed, so that the old and new classification tables were connected by reclassifying patents. Finally, combined with the revised classification and reclassification patents of Part H from 2009 to 2018, the preliminary exploration of technological evolution was carried out. [Result/conclusion] The patent reclassification model can effectively solve the problem of archived documents, provides reference for patent reclassification work, and can link the new and old patent classification tables; based on IPC classification revision and reclassification of patents, the main technological evolution direction after classification revision can be analyzed, providing a new perspective for research on technological evolution.
[1] 谢寿峰. 基于专利分析的技术演变与预测研究[D].南京:南京理工大学,2014.
[2] 刘云, 刘璐, 闫哲. 基于专利计量的全球碳纳米管领域技术创新特征分析[J]. 科研管理, 2016(S1):337-345.
[3] SUZUKI J, KODAMA F, et al. Technological diversity of persistent innovators in Japan:two case studies of large Japanese firms[J]. Research policy, 2004, 33(3):531-549.
[4] 廖列法, 勒孚刚. 基于LDA模型和分类号的专利技术演化研究[J]. 现代情报, 2017(5):15-20.
[5] 陈伟,林超然,李金秋,等.基于LDA-HMM的专利技术主题演化趋势分析——以船用柴油机技术为例[J].情报学报,2018,37(7):732-741.
[6] YOON J, KIM K. Identifying rapidly evolving technological trends for R&D planning using SAO-based semantic patent networks[J]. Scientometrics, 2011, 88(1):213-228.
[7] SASAKI H, KAJIKAWA Y, SAKATA I, et al. Predicting the potential industrial fields of technological spin-offs by using IPC in patent analysis[C]//Proceedings of technology management for emerging technologies.Vancouver:IEEE, 2012:975-991.
[8] ZHENG J, CHEN D Z, HUANG M H, et al. Industry and technology development in China from 2003 to 2008:a perspective from patent classification analysis[C]//Technology management for global economic growth. Phuket:IEEE, 2010:2256-2263.
[9] ZHOU X, ZHANG Y, PORTER A L, et al. A patent analysis method to trace technology evolutionary pathways[J]. Scientometrics, 2014, 100(3):705-721.
[10] JUN S. IPC code analysis of patent documents using association rules and maps-patent analysis of database technology[C]//International conference on database theory and application/international conference on bio-science and bio-technology. Jeju Island:SPRINGER-VERLAG BERLIN,2011:21-30.
[11] 黄斌, 黄鲁成, 吴菲菲,等. 基于专利共类的技术关联特征识别[J]. 情报杂志, 2015(1):44-48.
[12] CHEN J K C, PHAM V K, LIN F C, et al. Studying the patent of technology development on Dye Sensitized Solar Cell[C]/Technology management in the it-driven services. San Jose:IEEE, 2012:985-996.
[13] 李瑞茜, 陈向东. 基于专利共类的关键技术识别及技术发展模式研究[J]. 情报学报, 2018, 37(5):49-56.
[14] KRAFFT J, QUATRARO F, SAVIOTTI P P. The knowledge-base evolution in biotechnology:a social network analysis[J]. Economics of innovation and new technology, 2011, 20(5):445-475.
[15] 方曙,胡正银,庞弘燊,等.基于专利文献的技术演化分析方法研究[J].图书情报工作,2011,55(22):42-46.
[16] 吴红, 伊惠芳, 马永新, 等. 面向专利技术主题分析的WI-LDA模型研究[J]. 图书情报工作, 2018,62(17):68-74.
[17] 雷滔,陈向东.ICT领域跨学科关联度的块段模型分析——基于我国校企合作申请的专利[J].中国软科学,2011(1):67-74.
[18] 苗红,王晓宇,黄鲁成,等.基于专利的跨领域技术特征测度与分析[J].情报学报,2015,34(5):484-492.
[19] LAFOND F, KIM D. Long-run dynamics of the U.S. patent classification system[J]. Journal of evolutionary economics, 201,29(2):631-664.
[20] EUSEBI C A, SILBERGLITT R. Identification and analysis of technology emergence using patent classification[R]. California:RAND Corporation, 2014.
[21] WANG C C, SUNG H Y, HUANG M H. Technological evolution seen from the USPC reclassifications[J]. Scientometrics, 2016, 107(2):537-553.
[22] 王文晶. 新能源汽车产业专利分类修订与产业发展关联初探[J]. 科技创新与应用, 2017(14):25-26.
[23] 牛力, 胥凯, 王扬平. 空调领域IPC的修订及在检索中的应用[J]. 中国发明与专利, 2012(S1):136-139.
[24] 世界知识产权组织(WIPO)[EB/OL].[2020-11-14].. http://www.wipo.int/classifications/ipc/zh/.
[25] 世界知识产权组织.Guidelines for revision of the IPC[EB/OL].[2020-11-14]. https://www.wipo.int/export/sites/www/classifications/ipc/en/general/guidelines_revision_ipc.pdf.
[26] MIKOLOV T, CHEN K, CORRADO G, et al. Efficient estimation of word representations in vector space[J].arXiv preprint. arXiv:1301.3781, 2013.
[27] KIM Y. Convolutional neural networks for sentence classification[C]//Proceedings of the 2014 conference on empirical methods in natural language processing. Doha:Association for Computational Linguistics, 2014:1746-1751.
