语义相似度计算在语义标注中的应用

摘要

This paper presents a methodology to calculate semantic similarity between words and ontology concepts. This new approach uses the edit distance and the wikipedia to measure the semantic similarity between words and ontology concepts in terms of syntax and semantic. And under the guidance of domain ontology, it establishes mapping relations between words and ontology concept in the process of semantic annotation. At the same time, building a knowledge base to improve algorithm performance, experimental results show that this method is significantly effective.%提出了一种词汇和本体概念间的语义相似度计算方法.该方法利用编辑距离和维基百科从语法和语义两方面综合考虑词汇和概念间的语义相似度.在领域本体的指导下,将方法应用于语义标注过程,建立词汇与本体概念之间的映射.在标注过程中建立知识库,提高算法性能,实验结果说明该方法是行之有效的.