一种基于背景知识的本体注释方法

Approach of Ontology Annotation Based on Background Knowledge

本体采用基于语法词汇的表述方式,使本体自身表示可能存在模糊性、错误理解等问题,部分本体的概念可以通过自身的上下文信息推测出其含义,但是有些本体根据已有信息不能清晰表达其概念的确切含义.针对这个问题,提出基于背景知识的本体注释方法,对本体本身进行注释和澄清.包括基于WordNet和Web搜索引擎的注释方法,利用WordNet查找本体概念的正确词义,利用Web搜索引擎搜索本体概念的snippets,分别将词义和snippets作为其属性注释到本体中.实验表明本体注释率达到99.12%,表明本文方法的是可行的,本体注释正确率达到80.76%,比同类方法更高.

The lexical-based representation of ontology lead to the ambiguity and misunderstanding,we may inference the meaning of concepts by their context in some Ontologies,but some can not.To solve this problem,propose an approach for ontology annotation based on background knowledge.It consists of annotation approach and based on WordNet and approach based on Web search engine,search the right senses for the concept name,search the right snippets for the concept name,and annotate the senses and snippets into Ontologies as new properties.Experiment shows annotate ratio is 99.12%,which indicates the applicable of our annotation approach correct annotate ratio is 80.76%,does better than related approaches.