摘要
常规的学科热点及趋势监测方法存在监测成本高、监测信息相对滞后等问题。因此需要寻求较低成本、较快速的方法进行学科热点和趋势的监测，以保证监测系统提供信息的时效性。本文以图书情报档案学科为例，依据学科学术论文全文，从主题角度对学术论文集合进行全面分析，即：采用主题聚类方法，对包括时间信息的学科学术全文进行主题分析与主题聚类，归纳出某一特定学科的研究热点和这些热点的发展趋势。实验结果表明，基于主题聚类的学科热点及其趋势监测方法，其监测结果在很大程度上接近于常规方法的监测结果，但基于主题聚类的监测方法，在监测成本和监测信息时效方面得到改善。
Detecting hotspot and trend of discipline based on the traditional method is costly, time-consuming and hysteretic. It is urgent to design a system to detect hotspot and trend of discipline rapidly with low cost. The authors take the library, information and documentation science as an example, and have a systemic and comprehensive analysis on the academic papers collection with full-text according to the topic of papers. The method based on topic clustering is proposed and used to analysis and cluster the academic paper including the publication date. This method can detect hotspot and trend of a particular discipline. Experimental results show that the detecting results using topic clustering, to a great extent, are similar to that of the traditional methods in the task of detecting hotspot and trend of discipline. According to the cost and time, the detecting method based on topic clustering can improve performance of detecting system.
出处
《情报学报》
CSSCI
北大核心
2010年第2期342-349,共8页
Journal of the China Society for Scientific and Technical Information
基金
本研究受“十一五”国家科技支撑计划重点项目（2006BAH03B02）
中国博士后科学基金资助项目（20080430463）
南京理工大学科研启动基金项目（AB41123）资助.
关键词
学科热点监测
主题聚类
主题抽取
文本聚类
detecting hotspot of discipline, topic clustering, topic extraction, document clustering
