Research On The Concept Semantic Similarity And Correlation Based On Ontology And Application

Ontology as a semantic description tools have been widely used in many fields, whichdirectly leads to the expansion of domain ontology. This affected the efficiency betweenknowledge sharing and reuse. Being face of massive information, people hope to makescientific research and enterprise management on the foundation of the analysis of data, so as tobring economic and social benefit.The degree of similarity and correlation is a correlation between of any two concepts.Similarity and correlation between of any two concepts is the basis of ontology mapping,semantic retrieval, data mining, knowledge management, word disambiguation, semantic Weband other areas. With the increasing of the size of ontology, the exchanging of informationbetween ontology has become an important problem to be solved now. Calculating thesimilarity and correlation between ontology is a basic work on information exchange andprocessing. How to integrate and make full use of the existing ontology resources is needed forconcept semantic similarity and related computation support.The main content of this paper is divided into two aspects: one is the calculation method ofconcept semantic similarity and correlation based on domain ontology.Two is the application ofthe algorithm, construction of three dimensional semantic retrieval model to realize designknowledge reuse and reduce the loss of manpower and material resources. The main researchcontents are as follows:The way to calculate the semantic similarity of concept based on semantic distance,through the feature attribute to compute semantic distance, through the hierarchy depthcorrected similarity. In order to make full use of structural information in the ontology, thealgorithm combines the compute of correlation, so the result is more consistent with the actual.Because of the ontology designer’s knowledge background or different applications, thereare different levels of conflict between ontologies. In order to achieve the communicationbetween different ontologies, this paper presents a calculation method of concept semanticsimilarity calculation between different ontologies. This paper take the similarity of conceptname, the similarity of attribute and the similarity of structural into consider. The similarity ofstructural not only parent-child hierarchical relationship to be considered, but also increased theimpact of sibling similarity. Fusion three kinds of calculation methods, we can get the conceptsemantic similarity and validated by experiment.Domain ontology is a description of specific knowledge and information of the domain.According to the semantic knowledge is applied to the field of information retrieval, this paperdesign a simple model of semantic retrieval based on ontology, through the effectiveness ofretrieval results verify the proposed similarity and correlation algorithm. Used thethree-dimensional modeling design knowledge ontology as the experimental data, through theway of similarity and correlation calculate the similarity of the search condition and ontologylibrary. To establish a correspondence between the search condition and the three-dimensional modeling design knowledge ontology, help designers to find instance information and designknowledge, realize the reuse of design knowledge, reduce duplication of work of the design,raise the design efficiency and cut the product design cycle down.