Domain Ontology Construction Method Based On Relational Database

As a system model and modeling tool which could describe Information systemon both of semantic and knowledge layers, ontology has been widely used andeffective in the Knowledge Engineering, Artificial Intelligence and semantic webrecently. However, as the basis for ontology applications, ontology construction isvery complex, the traditional manual method of constructing the body is only a smallproportion of the ontology experts to master. The traditional course of constructingontology library requiring a lot of manual searches, input, instance, etc., is complexand hard-working. The building time depends on the level of the complexity andwhether ontology experts understanding that field. Ontology Construction bytraditional method is often overlook the details of the relationship of the domainconcept and easy to lead to knowledge acquisition bottleneck.In reality, relational database is already quite mature in knowledge representation,information systems related areas have the background as a relational databasesupport. Due to a lot of preliminary relational database works, the conceptual modelhas been relatively completed. Therefore, we can use the existing relational databaseresources to extract relevant concepts and relationships to construct ontology,relational data schema in fact contains a conceptual model related fields. So therelational database can sublimate from data management to knowledge management.This essay is the research about the domain ontology construction method basedon relational database, the approach avoids a large number of preliminary manualoperation. The paper puts forward with a building program with two conversions, atfirst, it extracts conceptual model from relational database resources, then we makethe conceptual model into the OWL ontology. First, we take relational database as adata source, in-depth analysis of several semantic information existing in relationaldatabase, then we makes use of the opposite project instrument to obtain the enddatabase of the conceptual data model. The essay explores the correspondingrelationship between ER model and OWL ontology in order to define thecorresponding mapping rules, including concept mapping rules, property mappingrules, the concept of the relationship mapping rules, axioms mapping rules andinstances mapping rules, respectively generating ontology concepts, attributes, levels,axioms and instances. Last, in this essay, we take school student management database system as a sample, verifying the actual feasibility of this method. Theresults showed that the method decreased a quantity of handwork, improving themodel construction proficiency as well as that ontology quality,which is much betterthan the traditional one.