An Improved Method For Ontology Matching

With the rapid development of the Internet of Things and the big data, thedemand for intelligent devices and intelligent search become more and moreincreasing, related research become more and more profound, Semantic Web is one ofthe important research branches. Ontology as the core of the Semantic Web, in recentyears, has been paid attention by many research institutions and organizations, obtaina number of related research achievement. In the Semantic Web systems, ontology hasdistributed nature, due to ontology construction has not yet formed a uniform standard,so different organizations in building ontology will cause the phenomenon ofontology heterogeneity, resulting a large number of Semantic Web informationsharing and interaction problems. Ontology matching is an effective way to solve theproblem of ontology heterogeneity. It decide the semantic relationships betweenentities by similarity value that obtained by calculated and analyzed the similaritybetween the ontology entity, in order to achieve semantic ontology compatible, thishas important significance for more effective sharing and reusing ontology inSemantic Web applications.In response to these questions about the ontology matching, this paper based on alot of literatures has done so many works. The main contents are listed as follows:Firstly, the paper introduced the concept of ontology and ontology descriptionlanguage, on this basis, introduced the concept and process of ontology matching thatis the most effective way to solve the problem of ontology heterogeneity in detail,finally introduced several ontology matching algorithm and domestic and foreigntypical matching tool. Secondly, by reading a lot of literatures and analyze manyexisting matching method. This paper proposed an improved method for ontologymatching. The specific process is as follows: generate RDF graph by ontology, thengenerate a tree by RDF graph with some rules, traversal the tree from top to downaccording to the order, calculate semantic similarity and structural similarity of eachnode, then determine the two entities has matched or not by similarity value, get the set of matching result. Thirdly, through experiments to verify the improved methodproposed in this paper. This paper adopts Java language and uses Jena API to designthe experiment, uses benchmark data set of Ontology Alignment Evaluation Initiative2010as experiment data, verify the effectiveness of the proposed algorithm byprecision, recall, F-Measure value. By comparing the experimental results betweenthe improved method proposed in this paper and GMO of Falcon-AO system, weanalysis the advantages and disadvantages of the improved method proposed in thispaper. Experiments results prove that the method is effective.