Semantics-driven ontology clustering
Ding Qiu-lin · Ha'erbin gongye daxue xuebao · 2008
An ontology similarity computing method based on semantic is proposed. This method considers the similarity of the concept itself as well as the similarity of the attribute set and the related concept set. Basic similarity threshold of the concept was used to control the probability to compute the similarity of the attribute sets, and semantic radius was used to control the scope of related concepts. The ontology similarity method based on semantics and an agglomerative hierarchical clustering algorithm were combined to realize the ontology clustering driven by semantics. The experiment shows that the ontology clustering based on semantics can get satisfying results and its performance outperforms the ontology clustering using oMAP similarity method. The proposed method has been applied in a certain aeronautical institute and plays a special role in ontology integration.