基于数据挖掘的产业技术演化路径识别方法研究  

Research on the identification method of industrial technology evolution path based on data mining

作　　者：和志强[1] 王梦雪 马宁 刘院英[1] HE Zhi-qiang;WANG Meng-xue;MA Ning;LIU Yuan-ying(School of Information Technology,Hebei University of Economics and Business,Shijiazhuang Hebei 050061,China)

机构地区：[1]河北经贸大学信息技术学院,河北石家庄050061

出　　处：《河北省科学院学报》2022年第4期8-13,共6页Journal of The Hebei Academy of Sciences

基　　金：河北省省级科技计划资助项目(19210105D)。

摘　　要：开展产业技术演化路径的识别与分析,对整体把握产业技术发展态势、促进区域经济发展具有重要的现实意义。为识别产业技术演化路径,提出了一种基于专利文献数据挖掘的技术演化路径识别方法。该方法采用主题模型提取技术主题与技术关键词,同时抽取各技术主题的代表性专利。并在此基础上,利用专利间相似度来发现专利间关联关系实现技术演化路径识别。通过选取塑料包装产业相关专利进行实证研究,表明该方法能够有效提取产业相关技术的代表性技术并识别其演化路径。Identifying and analyzing the evolution path of industrial technology is important for grasping the development trend of industrial technology as a whole and promoting regional economic development.In order to identify the evolution path of industrial technology,a technology evolution path identification method based on patent documents data mining is proposed.The method uses the topic model to extract technical topics and technical keywords,and simultaneously extracts representative patents of each techni-cal topic.On this basis,the similarity between patents is used to find the correlation among patents to re-alize the identification of technology evolution path.Empirical research by selecting patents related to the plastic packaging industry shows that the method can effectively extract the representative technologies of related technologies in the industry and identify their evolution paths.

关 键 词：技术演化 主题抽取 TextRank 关键词提取 关联关系抽取 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]