国内外专利挖掘研究(2005-2014)综述

在中国知网、万方数据和Web of Science进行检索,获得72篇相关中文文献和98篇英文文献,并从中选出66篇进行综述。专利挖掘研究包括术语抽取、聚类、分类、以复杂网络为基础的方法、以时间为基础的方法和基于专利挖掘的技术研究等6个主题。尽管近10年来这一领域发展较快,但是部分研究也存在试验验证不精确、基于IPC的自动分类效果不好、所要解决的问题不明确且局限于方法应用和粒度粗糙等问题。专利挖掘研究应该注重发现问题,而非简单地应用方法。

A Review of Patent Mining Studies in China and Abroad 2005-2014

The paper performs an exhausted literature review on patent mining. The authors retrieve CNKI, Wanfangdata and Web of Science, and gets 72 Chinese and 98 English papers. The review carefully chooses 66 from them. Patent mining includes the following topics: term extraction, clustering, categorization, methods based on complex network, methods based on time and technical studies based on patent mining. Although patent mining develops fast in last 10 years, there are problems calling for improvements. For example, test experiment lacks accuracy; the effectiveness of automatic categorization based on IPC is not satisfactory; the research question is not clear and certain research is contented with the application of a given method; the granularity is coarse, etc. The paper concludes that patent mining should pay adequate attention to putting forward new questions rather than merely applying existed methods to patents.