Recognizing the Relations between Web Pages Using Artificial Neural Network

Semantic Web shows us the potential infrastructure of the next generation Web. Web information will be understandable to machines in this infrastructure. Our research work 1 has the aim of embedding machine-understandable semantic information in ordinary HTML files automatically given the domain ontology. We focus on automatically acquiring the instances of concepts and relations defined in domain ontology from Web pages. This paper describes how to recognize the relation between Web pages (a kind of relation instance) by using Artificial Neural Network (ANN). The input vector of ANN is determined by the type of Web pages, the number and type of hyperlinks between the Web pages, and the similarity in Web pages ’ contents. We also show the initial results got by our prototype system