中文摘要：

传统的超文本标记语言只能显示而不能使计算机理解页面信息,导致传统的基于关键词的检索工具的查准率、查全率和智能化程度都较低。虽然语义网可以解决计算机对页面信息的理解问题,但以往基于本体的语义匹配算法通常是面向类概念的,在精度上不能满足微内容的发现需求。本文面向微内容检索应用,在分析了个体概念的语义特征的基础上,提出了基于语义关系的个体匹配规则、索引结构和相应的语义检索算法。最后,通过实验验证,证明了基于个体语义关系的聚合机制具有更高的聚合细腻度,所提出的语义检索算法在面向微内容的检索应用中是有效的。

英文摘要：

Traditional HTML can only display but not make computer understand the information on pages,leading to low precision recall rate and intelligence of traditional keyword-based search tools.Based on Semantic Web,computer can understand the information on pages,but previous ontology-based semantic matching algorithm which is usually based on category concept can not meet the needs of micro-content retrieval in accuracy.In this paper,for micro-content retrieval applications,an analysis on the semantic features of individual concept is done first.On this basis,the individual-matching rules and indexing structure based on semantic relations and then the corresponding semantic retrieval algorithms are presented.Finally,basis on the experiment,it＇s proved that the index which is designed based on semantic relations have a higher degree of polymerization,and the retrieval algorithms is effective for the application of micro-content retrieval.