文档详情

基于本体的信息检索模型研究.pdf

发布:2017-05-10约2.24万字共9页下载文档
文本预览下载声明
() 41  2 Vol.41, No.2    JOU RNAL OF NANJING UNIVERSITY      2005 3 Mar., 2005 (NATU RAL SCIENCES) 宋峻峰 , 张维明, 肖卫东, 唐九阳 (, , 4 10073)  : , , , ., , tablea , , ;, , . : ,  ,  , : TP 391 Research on Ontology Based Information Retrieval Model Song J un _Feng, Zhang Wei _Ming, X iao Wei_Dong , Tang Jiu _Yang (School of Information System and Management, The National University of Defense Technology, Changsha, 410073, China) Abstract: In the traditional information retrieval models, altho gh s itable retrieval f nction is chosen in terms of logical views, the performance of information retrieval is always nsatisfiable beca se the logical view s of doc ments and the logical views of sers information needs can not represent doc ments and ser s information needs well.In this paper, ontology_based information retrieval model is introd ced;ontology is generated sing a kind of basic description logic, w hich is a s itable tradeoff between expressivity of know ledge and complexity of reasoning problems.The eq ivalent classes of concepts can be obtained by sing tablea algorithm for this kind of basic description logic;the eq ivalent classes of individ als can be obtained by sing eq ivalent relations between individ als nder condition that only atomic roles exist;so the set of semantic index terms can be acq ired.The logical views of doc ments and ser s information needs, generated in terms of these semantic index terms, can represent doc ments and ser s informati
显示全部
相似文档