Knowledge is an essential part of most Semantic Web applications and ontology. Ontology is the most important part of the knowledge. Ontology learning, which refers to extracting conceptual knowledge from several sources and building ontology from scratch, enriching, or adapting an existing ontology is one of the attempts at knowledge acquisition. Ontology is not sufficient to represent inferential knowledge. The ontologybased analysis with description logic is a popular issue of the Semantic Web. Rules are obtained from several sites of the same domain. The existing system has some problems in rule extracting. First, the web pages are identified for rule components and for their types. Second thing is how to compose the rules with rule components. The domain has similar Web sites explaining similar rules from each other. It decreases the burden on the knowledge experts and domain experts. Our idea for solving these problems is using rules of similar sites in limited situations. The two main steps of rule acquisition, which consists of rule component identification such as variables and values in Web pages by using RuleToOnto and rule composition with the identified rule components. We performed experiments demonstrating that our ontologybased rule acquisition approach works in a realworld application.
展开▼