受限本体相似

在从不同的语义Web上得到用本体表达的文档资源以后,这些文档资源通常被转换成基于同一个本体的本体描述,这样既便于对文档的分析,又便于在此基础上进行信息抽取.这些文档本体之间仅仅在实例和关系层上彼此相互不同,在类、属性、规则、谓词方面都基本相同.对这种文档的检索,一个最普通的操作就是计算本体之间的相似性.很多计算本体相似性的方法基本上都是以分别属于不同本体的实体之间配对比较来实现,而且往往要考虑所有相关的元素.这不仅增加了计算复杂度,还会遇到循环计算的问题.在对语义网本体语言的推理能力进行研究以后,提出了一种基于知识推理的二阶本体相似技术,解决了循环计算的问题.

Restricted Ontology Similarity

Resources integrated from heterogeneous resources of the Semantic Web are usually transformed to ontologies which based on the same ontology schema and can be used for analysis and data mining. These ontologies are different from each others in the part of instances of classes and properties. A variety of methods can be used for the similarity of these ontologies. Basically they apply pair-wise comparison of entities from each of the ontologies and select the most similar pair. These entities include class, property, instance, data type and so on. However, this task can be restricted in the comparison of instances and relations. A similarity measure is introduced for the similarity of restricted ontologies, which makes use of the specialty of inference in ontology and is a 2-grade ontology similarity technique. Moreover, the 2-grade ontology similarity technique can resolve the problem of circularities in similarity definition.