摘要
由于跨领域专利具有跨越学科边界,融合多个领域的理论特征,且技术主题范围不明确,关键词涵盖范围大,目前传统的专利检索方法并不适用。本文在传统专利检索方法的基础上主要通过专利数据库选择、专利IPC分类号位置识别、关键词的确定和基于机器学习的专利文本识别与分类等四个步骤,并结合专家智慧实现对跨领域专利的检索,提出一种适用于跨领域专利的检索策略。同时,在“金融科技”领域进行实证研究与分析,证明了该策略的有效性,也为跨领域专利检索工作的开展提供借鉴。
Because cross-disciplinary patents have the theoretical characteristics of crossing disciplinary boundaries and integrating multiple fields,and the scope of technical topics is not clear,and the scope of keywords is large,the current traditional patent search methods are not applicable.Based on the traditional patent search method,this article mainly adopts the four steps of patent database selection,patent IPC classification number position identification,keyword determination,and machine learning-based patent text recognition and classification,and combines expert wisdom to achieve cross-disciplinary patents search.Moreover,we propose a search strategy which are suitable for cross-domain patents.Meanwhile,empirical research and analysis in the field of“Fintech”proved the effectiveness of this strategy and provided reference for the development of crossdisciplinary patent search.
作者
高辰琛
刘琦岩
望俊成
张玄玄
GAO Chenchen;LIU Qiyan;WANG Juncheng;ZHANG Xuanxuan(Institute of Scientific and Technical Information of China,Beijing 100038,China)
出处
《情报工程》
2020年第5期90-99,共10页
Technology Intelligence Engineering
基金
中国科学技术信息研究所重点工作项目“金融大数据建设与知识服务(二期)-金融科技知识图谱构建”(ZD2020-03)。
关键词
跨领域专利
检索策略
IPC
关键词
文本分类
Cross-disciplinary patent
search strategy
IPC
key words
text categorization
