Web oriented learning strategies of relationships among domain ontology concepts

In order to learn the taxonomic and non-taxonomic relationships among ontology to help knowledge engineers construct domain ontology, we propose a method to obtain relationships among concepts automatically in Chinese domain ontology learning environment. The method is based on Web data to extract candidate relations and discusses relationships extraction based on webpage link, categories, free text or non-taxonomic relationships according to different relationships. It presents the learning strategies of relationships among domain ontology from web under Chinese environment and explores the methods of learning concepts relations from multiple angles, which can overcome the one-sidedness of a single method effectively.