摘要
针对传统信息检索模型不能很好满足用户需求的问题,在分析现有相关研究的基础上,提出基于领域On-tology的知识检索模型。通过构建领域Ontology,对文档进行语义标注,对查询请求进行概念提取和语义扩展,从而得到语义索引项作为文档和用户请求的知识表达,进一步研究领域Ontology中词语间语义关系的计算模型。考虑到语义相似度与语义相关的内在关系,给出相关系数来衡量检索目标与候选者间符合程度。最后对提出的模型进行验证,结果表明检索性能有显著提高。
Traditional information retrieval model becomes inefficient for users. In this paper, the knowledge retrieval model based on domain ontology is firstly proposed by analyzing the existing research results. By constructing domain ontology, the documents are semantic annotated, the queries are conceptualized and semantic extended, and semantic indexing terms are generated, which representing documents and queries. And then, it introduces a computing model of words in domain ontology, which takes full account of the semantic similarity and relevancy. Finally, according to the application scenario, the performance of retrieval can be improved effectively, which shows advantages over the keyword-based information retrieval.
出处
《图书情报工作》
CSSCI
北大核心
2010年第18期116-119,144,共5页
Library and Information Service
