Ontology Matching And Repair Method Based On Semantic Association

With the continuous development of semantic Web technology,computers can better understand the massive knowledge information in the Web,and people can search the knowledge content they need more conveniently.Since knowledge is represented in various structural or unstructured forms,the realization of knowledge interoperability can help us relate more relevant knowledge to provide high-quality search services.In recent years,more and more researchers use ontology technology to construct domain ontology,but because there is no unified standard for the construction,there are different ontology model structures for the same domain knowledge.If the differences can be explored,the ontology-based knowledge framework can be improved.In heterogeneous ontology,two concepts often refer to the same thing,and it is of great significance to integrate their mapping together.It is very difficult to find semantic association between concepts when there is a large difference between ontologies.In addition,logical conflicts are easy to occur in the process of ontology matching and domain ontology construction.For ontology matching,the matching repair method can not only deal with logical errors but also detect the most likely false matching pairs.If the matching method is applied to the real world in the future,the results must be tested and repaired.In addition,a complete domain ontology can deduce hidden semantic information based on existing knowledge,so it is necessary to ensure the accuracy of domain ontology logic.Logical diagnosis of ontology is a very important task and is also one of the steps of ontology repair.Therefore,the research content of this paper mainly focuses on how to efficiently match and repair ontology,as follows:(1)Ontology matching based on semantic association and probabilistic logic is studied.This paper mainly proposes an ontology matching method based on semantic association,which uses the existing semantic knowledge to calculate semantic similarity and screen out accurate matching pairs,and then calculates the structural similarity between concepts according to the characteristics of the context structure.Most of the correct matching pairs can be found through the two similarity calculation methods.Therefore,the matching method can carry out the matching work better on the basis of reliable knowledge.In addition,this paper introduces a probabilistic logic repair method,which can detect and repair the conflict of matching results.At the same time,the string length of different ontology is limited to improve the efficiency of matching repair.(2)Ontology repair based on semantic association and description logic is studied.In this paper,a selection function algorithm based on dynamic cache window is proposed,which constructs relevant axioms from semantic association among axioms,selects a certain number of axioms based on the window size,and then calls inference machine for diagnosis.In this algorithm,the size of cache window is dynamically adjusted according to the change of concept unsatisfaction.This will improve the tedious process of manually adjusting different ontologies to get the optimal number of Windows in the fixed cache window method.In addition,the algorithm minimizes the time consumed by ontology diagnosis.