基于文本预处理的德温特专利信息分类方法研究  

Research on the Classification of Derwent Patent Information Based on Text Pretreatment

作　　者：赵蕴华[1] 张静[1] 崔伟[1] 

机构地区：[1]中国科学技术信息研究所,北京100038

出　　处：《情报科学》2012年第10期1452-1455,共4页Information Science

摘　　要：针对原始德温特专利信息内容表示粒度大,不能够满足对该数据直接进行专利统计分析的问题,本文提出了一种基于文本预处理的德温特专利信息分类方法。该方法结合文本预处理技术,通过对专利数据结构和内容的分析,实现了对专利内容进一步分类。实验结果表明,该方法有效的降低专利信息内容的表示粒度,从而提高了专利分析结果的准确性和多样性。In order to solve the problem that the large-grained expression of the original derwent patent information can not be directly analyzed, a new method of the classification of derwent patent information based on text pretreatment is proposed. This method analyzes the structure and content of the patent information and implements further classification by combining text pretreatment. The experimental results demonstrate that the method reduces the grain effectively, and improves the accuracy and diversity of analytical results.

关 键 词：德温特创新索引 文本预处理 专利信息分类 专利分析 

分 类 号：G254[文化科学—图书馆学]