一种基于领域本体的论文检索方法的研究与应用

摘要

Aiming at the shortcomings of traditional dissertation retrieval approaches in lack of semantic understanding and low retrieval results pertinence,in this paper we present a domain ontology-based retrieving method using document semantic presentation model based on semantic network.Firstly,it builds the domain ontology in combination with scientific classification system,and makes semantic indexing on dissertation documents.Secondly,according to ontology knowledge and index information,it constructs semantic networks-based expression model of documents semantics.Finally it modifies users query and the correlation algorithm of semantic networks,integrates the keyword-based and semantic-based approaches to implement results sorting.Experimental results show that the proposed method can improve the accuracy and call rate of dissertation retrieval effectively.%针对传统的论文检索方法缺乏语义理解，检索结果相关度不高的缺点，采用基于语义网络的文档语义表达模型，提出一种基于领域本体的检索方法。首先结合学科分类体系构建领域本体，并对论文文档进行语义索引；然后根据本体知识和索引信息构建基于语义网络的文档语义表达模型；最后改进用户查询与语义网络的相关度算法，综合关键词和语义的方法实现结果排序。实验结果表明，该方法能有效地提高论文检索的准确率和召回率。