Study Of Ontology-based Web Mining Classification Method

With the rapid growth of the Web, the Web information has became the socialcommonality information resource. With the scale of Web information growsexplosively, how to analyze, explore and discover useful knowledge rapidly andefficiently from them becomes the focus of researchers.This thesis represents a new classification method based on the ontology,asfollows is the main works in this thesis: Designing the new classification method after analyzing the Web data miningand semantic ontology. This method is composed of user session level anddomain level, based on the level of user sessions data mining, this thesismake the most of the potential semantic of the domain ontology creating theusage profile of a set of structured Web objects. At the user session level, designing and implementing the FCM algorithm,improved the data input;clustering number, weighting exponent and creatingthe user profile show the data mining results. At the domain level, implementing the objects of domain ontology instanceand the semantic of the classification result based on the Web mining resultand ontology application. The example shows how creating combinationfunction for the class attribute, implementing the domain level mining, showsthe reason why these users are grouped together. As an application framework, creating an instance model show theapplication of the new classification method.Above all, this thesis implements a new Web mining classification method whichbased on the ontology, setting up a using model and validating in theory, explainingby the example. This method provides a powerful tool for the Web data mining, theusing of Web resource and the Web personalization services.