基于词汇链的应急预案主题抽取方法研究  

Research on Semantic Relatedness Based Subjects Extraction from Emergency Plans Literature

作　　者：裘江南[1] 罗志成[2] 王延章[1] 

机构地区：[1]大连理工大学管理学院,大连116024 [2]武汉大学信息资源研究中心,武汉430072

出　　处：《情报学报》2008年第6期891-896,共6页Journal of the China Society for Scientific and Technical Information

基　　金：国家社会科学基金资助项目（批准号：07CTQ006）

摘　　要：本文针对应急预案自动主题抽取的需求，采用词汇语义相关度计算的方法，构建了一个基于词汇链算法且符合人的主观感受的主题抽取模型。模型根据应急预案文本的特点，运用了自然语言处理技术，改进了原始的词汇链生成算法，提出了一种多因素词语权重算法。通过与人工抽取主题词的实验结果相比较，该主题提取模型在查全率和查准率上都取得了较好的效果。The paper aimed at the requirement of the automatic extraction of subject from the emergency plans, took up with the measures of lexical semantic relatedness, and has constructed a subject extraction model based on the lexical chain algorithm which accords with human＇ s subjective feeling. According to the characteristics of the emergence plans text and the needs of the project, the model used a number of natural language processing methods, improved the original chain generating algorithm, and brought forward a weight algorithm based on multi-factors. Finally, an experiment was carried out which compared the human subject extraction results to our system result, and the recall and the precision showed that our model do a good job.

关 键 词：主题抽取 词汇链 语义相关 应急预案 

分 类 号：G250.73[文化科学—图书馆学] TP391.4[自动化与计算机技术—计算机应用技术]