Title :
Ontology mapping based on structures and instances
Author :
Zheng, Cheng ; Shen, Ya-Peng ; Mei, Lin
Author_Institution :
Key Lab. of Intell. Comput. & Signal Process., Anhui Univ., Hefei, China
Abstract :
Ontology mapping is to find the mapping relationship between different ontology to reuse and share ontology. In this paper we present a multi-strategy ontology mapping technique based on the structure and instance. When the similarity between the mapping pairs is approximating to the threshold, context structures are used to judge their similarity. The key of the instance mapping is calculating the instance similarity. Firstly the instance is preprocessed. Then set similarity and probability similarity based on Bayesian theory is calculated respectively. Finally instance similarity is calculated by the set similarity and probability similarity comprehensively. Experimental results show its effectiveness.
Keywords :
Bayes methods; ontologies (artificial intelligence); Bayesian theory; context structures; multistrategy ontology mapping technique; probability similarity; Cybernetics; Encoding; Machine learning; Ontologies; Optimization; Probability; Semantics; Ontology Mapping; Ontology instance; Ontology structure; Similarity;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2010 International Conference on
Conference_Location :
Qingdao
Print_ISBN :
978-1-4244-6526-2
DOI :
10.1109/ICMLC.2010.5581019