摘要
本体技术作为一种能在语义和知识层次上描述概念体系的有效工具,给词语间相似度计算带来了新的机会。词语相似度的研究,是知识表示以及信息检索领域中的一个重要内容。本文利用本体来组织概念,计算概念之间的语义相似度,将语义相似度分成概念相似度和描述相似度,把概念相似度和描述相似度进行合并,生成最终的语义相似度。依据《中国分类主题词表》建立的计算机领域本体,验证了语义相似度计算方法的有效性。
The research about word similarity is very important in knowledge representation and irfformation retrieval. Semantic similarity is divided into concept similarity and description similarity. A calculation of semantic similarity between concepts in ontology is presented by the hierarchy of ontology concepts. To make the calculation of semantic similarity between concepts more similar to human＇s subjective judgment, semantic similarity is divided into concept similarity and description similarity; the final semantic similarity is produced by merging concept similarity and description similarity. The optimized parameter settings are gained through the experiment results.
出处
《情报科学》
CSSCI
北大核心
2012年第8期1238-1241,共4页
Information Science
基金
黑龙江省高等教育学会高等教育科学研究"十二五"规划重点课题(HGJXHB110334)
关键词
本体
语义相似
概念相似
关系相似
属性相似
ontology
semantic similarity
concept similarity
relation similarity
attribute similarity
