基于中医药领域本体的信息检索模型研究

Study of the Information Retrieval Model Based on Traditional Chinese Medicine Domain Ontology

针对传统基于关键词匹配的中医药信息检索存在查全率和查准率低下的缺陷,将本体与潜在语义索引相结合,提出一种基于中医药领域本体的语义信息检索模型。该模型基于本体概念扩展树构建相应的查询扩展方法和语义向量空间模型,将用户查询和文档集映射到同一潜在语义空间,通过计算查询向量与文档之间的相似度返回检索结果。着重阐述了该模型的体系结构、实现过程和关键技术,并对其实用性进行论证。

Aiming at the defect of low precision rate and low recall rate of traditional Chinese medicine （TCM） infor-mation retrieval based on keywords matching, we propose a semantic information retrieval model based on domain ontology of TCM by combining ontology with latent semantic indexing. Based on ontology concept-extended tree method of query expan-sion and semantic vector space, the model can map user queries and documents to the same latent semantic space, and re-turning retrieval results by calculating the similarity between the query vector and the document. In this paper, we focus onthe architecture, implementation process and key technologies of the model, and demonstrate its practicability.