首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于专利的技术机会识别:深度学习领域的案例分析
引用本文:杨辰,王楚涵,陶琬莹.基于专利的技术机会识别:深度学习领域的案例分析[J].科技管理研究,2021,41(12):172-176.
作者姓名:杨辰  王楚涵  陶琬莹
作者单位:深圳大学管理学院,广东深圳 518060
基金项目:国家自然科学基金资助项目“基于科研社交网络挖掘的专家组合推荐问题的研究”(71701134);广东省软科学项目“粤港澳大湾区背景下城市科技创新生态系统构建—以深圳市为例”(2019A101002075);深圳市哲学社会科学规划课题项目“基于智能语义本体的科学研究动态分析研究”(SZ2020D015)
摘    要:为及时有效地识别潜在技术机会,采用文本挖掘和异常值检测的方法,提出一种基于专利文本的技术机会识别方法.首先采用文本表示模型Doc2vec技术对专利摘要进行建模,以更深层表征文本语义信息;然后利用基于密度的离群值检测算法,识别出具有潜在技术机会的专利方向;最后以深度学习领域潜在技术识别为例,构建专利检索式并收集458条专利文献作为数据集.实证结果总结出4类主题共10个潜在的技术机会,验证了该基于专利的技术机会识别方法的有效性,可为企业相应技术应用、研发和创新提供参考.

关 键 词:专利技术  技术机会  技术识别  专利分析  文本挖掘  深度学习
收稿时间:2020/10/15 0:00:00
修稿时间:2021/6/19 0:00:00

Patent-based Technology Opportunity Identification: A Case Study Based on Deep Learning
Yang Chen,Wang Chuhan,Tao Wanying,Geng Shuang.Patent-based Technology Opportunity Identification: A Case Study Based on Deep Learning[J].Science and Technology Management Research,2021,41(12):172-176.
Authors:Yang Chen  Wang Chuhan  Tao Wanying  Geng Shuang
Abstract:To identify potential technology opportunities timely and effectively, this paper proposes a technology opportunity identification method based on patent text by using the methods of text mining and outlier detection. Firstly, the paper uses Doc2vec to model the patent abstract to represent semantic information of the text. Then, the paper uses a density-based outlier detection algorithm to identify patents with potential technology opportunities. Finally, the paper takes the identification of potential technologies in the field of deep learning as an example for case analysis, constructs a patent search query and 458 patent documents are collected as the dataset. The empirical results summarize four kinds of topics and ten potential technology opportunities, which verifying the effectiveness of this patent-based technical opportunity identification method, and can provide reference for the enterprise corresponding technology application, research and development and innovation.
Keywords:patent technology  technology opportunity  technology identification  patent analysis  text mining  deep learning
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《科技管理研究》浏览原始摘要信息
点击此处可从《科技管理研究》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号