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 only counts the concurrence of words, but ignores their conceptual relations, which leads to the poor precision in information retrieval. Conceptual Graph (CG) based intentional logic model rebuilds the conceptual relations among words to better reflect user requirements and hopefully raise the precision of retrieval. In the article the authors propose a Query CG guided filtering method for webpage search results. Given Query CG, the authors look for concepts and relations included in the Query CG and build a simplified snippet CG. If the snippet CG can cover the Query CG, its snippet is reserved; otherwise its snippet should be rejected.
展开▼
