需求概念图导引下的网页检索结果过滤  

WEBPAGE RETRIEVAL RESULT FILTERING GUIDED BY QUERY CONCEPTUAL GRAPH

作　　者：刘华晖[1] 陈玉泉[1] 

机构地区：[1]上海交通大学计算机科学与工程系,上海200240

出　　处：《计算机应用与软件》2012年第2期17-18,63,共3页Computer Applications and Software

基　　金：国家自然科学基金(60873135)

摘　　要：基于关键词的布尔模型将用户需求视为词汇集合,只考虑词汇共现,忽略了词汇间的概念关联,用于信息检索精确度较低。基于概念图的内涵模型重建词汇间的概念关联,更好地反映用户需求,有望提高检索精确度。提出一种在需求概念图导引下对网页检索结果进行过滤的方法:给定需求概念图,在摘要中寻找需求概念图包含的概念及其关联,构建简化的摘要概念图,如果简化的摘要概念图能够覆盖需求概念图,则保留该摘要,否则将该摘要滤除。Keyword based Boolean model treats user queries as word aggregations.It on ly counts the concurrence of words,but ignores the ir conceptual relations,wh ich leads to the poor prec ision in inform ation retrieval.Conceptual G raph（CG） based intentional logic model rebu ilds the conceptual relations among words to better reflect user requ irem ents and hopefu lly raise the prec ision of retrieval.In the artic le the authors propose a Query CG gu ided filtering m ethod for webpage search resu lts.G iven Query CG,the authors look for concepts and relations inc luded in the Query CG and bu ild a simp lified sn ippet CG.If the sn ippet CG can cover the Query CG,its sn ippet is reserved;otherw ise its sn ippet shou ld be rejected.

关 键 词：概念图模型 需求概念图 网页检索过滤 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]