基于共引分析及Fisher判别法的网站过滤及扩展研究  

Websites＇ Filtering and Extension Based on Co-Citation and Fisher Ruler

作　　者：袁毅[1] 吴宗翔[2] 

机构地区：[1]中国矿业大学文法学院,徐州221008 [2]中国矿业大学理学院,徐州221008

出　　处：《情报学报》2006年第5期606-612,共7页Journal of the China Society for Scientific and Technical Information

基　　金：江苏省哲学社会科学“十五”规划基金项目（编号：04XWB013）.

摘　　要：从大量动态无序的非结构化网络资源中发现主题或学科核心网站，是学科导航、个性化服务及信息推送的基础。但在核心网站评选过程中，存在两个问题，一是通过搜索引擎获得的相关网站集合中，有大量非相关网站，直接影响核心网站评选的效率；二是当前核心网站研究中均未考虑到网站的扩展问题。为此，本文将Web共引分析技术引入非相关网站过滤，提出利用Fisher判别法实现核心网站扩展，并给出了扩展可靠性检验方法。Search subject core Websites from a lot of dynamic out-of-order Web resources is foundation of subject navigations, individuation services and information push. However, in the process of core Websites evaluation and selection, there are two problems. At first,there are many no-relative Websites in the result of research. This decreases core Website evaluation efficiency. The second, core Websites＇ expand is not discussed in the past studies. In this study, Web co-citation analysis is used to filter non-relative Websites, fisher ruler is used to expand core Websites. Further more, the test method of expand＇ reliability is given.

关 键 词：Web资源评价 网站过滤 网站扩展 核心网站 

分 类 号：G250.73[文化科学—图书馆学]