An ontology-based Web information extraction approach

An approach supervised by ontology is proposed for Web information extraction after analyzing two types of methods based on wrapper and concept model. Using concepts and taxonomy relation between concepts provided by ontology, this method can locate the wanted information blocks in Web page quickly by judging if adjacent sub-trees which are included in HTML Tree are isomorphic. Furthermore, combining text's data-modes the method can filter information which are irrelevant to the wanted information and achieve higher accuracy of information extraction.