摘要: 针对概念语义相似度计算方法在信息检索中存在的漏检、误检等问题,提出一种基于上下文的计算本体内概念间语义相似度的模型。该模型分别从概念的父节点、子节点以及概念间路径权重3个角度进行计算,加权求和并求得语义相似度。实验结果表明,该模型的计算结果更加接近专家的经验值,为概念之间的语义关系提供了有效的量化。
关键词:
语义相似度,
上下文,
本体,
语义距离
Abstract: The calculation method of the concept semantic similarity exists some problems in information retrieval, such as omitting retrieval false retrieval and omitting retrieval. Aiming at these problems, a semantic similarity model based on context is brought forward, which calculates the concept similarity of parent ones, child ones and the path weighted in the ontology. It calculates the sum of their weight, and obtains the semantic similarity. Experimental results show that the model results are closer to the experience of experts, and the values provide an effective quantification of the semantic relationship between the concept.
Key words:
semantic similarity,
context,
ontology,
semantic distance
中图分类号:
吕林涛, 董迎. 基于上下文的概念语义相似度计算模型[J]. 计算机工程, 2010, 36(21): 59-61.
LV Lin-Chao, DONG Ying. Concept Semantic Similarity Computation Model Based on Context[J]. Computer Engineering, 2010, 36(21): 59-61.