Research On Concept Similarity Algorithm Based-on Ontology

Concept similarity is the basis of ontology similarity.In Semantic Web environment, ontology is created according to the different purposes and methods, so that the contents of ontology often overlap or associated, resulting in heterogeneous ontology.However,the current technologies of ontology cannot recognize and estimate the semantic between different ontologies in order to share resource, potential semantic of ontology must be mined to solve the resource sharing of heterogeneous ontology.But the mining of implied semantic information of ontology depends on the research of concept similarity.In order to improve the accuracy of concept similarity and complete the semantic information of concept similarity, one of the algorithms is proposed in this thesis–concept similarity algorithm based-on ontology (CSO algorithm).CSO algorithm based on T-L BACH algorithm ,on the one hand,according T-L BACH algorithm without considering the structural relationship between concepts that the accuracy of concept similarity is not high,to import the formal concept analysis(FCA)theory,formal context is created by means of the ordering relationship between different ontologies,then similarity computational model based on concept lattice is structured according to formal context,the model computes concept similarity according to the meet-irreducible elements;on the other hand, according T-L BACH algorithm cannot fully express the semantic information of ontology in calculating RDF figure structure similarity of concept, to import the attributes of the OWL operator,RDF triples describes the semantic information of concept accurately by the comparability of OWL operator,to improve the accuracy of RDF figure structure similarity,then describes the semantic of the concept fully.According to the thought of CSO algorithm, the prototype framework drawing of CSO algorithm is provided and the content of CSO algorithm is designed in this thesis. In this thesis,in order to verify the advantage of the algorithm proposed, the calculation system of concept similarity based on CSO algorithmthe is designed, the values of concept similarity is calculated by T-L BACH algorithm and CSO algorithm, the values of concept similarity is compared by the chart.The results of the experiment shows that the values of concept similarity is consistent for CSO algorithm and T-L BACH algorithm in the general trend,but CSO algorithm is superior to T-L BACH algorithm in the accuracy of algorithm,so that CSO algorithm can reflect the semantic of the concept of ontology, lay the first stone to solve the resource sharing of heterogeneous ontology.At last,the thesis briefly introduces the application of CSO algorithm in the domain of information retrieval, so CSO algorithm make sense for research of information retrieval.