Ontology-based Information Filtering Technology Research

Traditional information filtering techniques, which take the interests of users into account only, fail to satisfy user requirements. In this thesis, an information filtering based on ontology is introduced. Users of filtering are classified two types: (1) no experiential users of filtering, (2) experiential users of filtering. Based on user profile, topic template is brought forward for no experiential users of filtering and additive user profile is brought forward for experiential users of filtering. Information is filtered based on user profile and topic template for no experiential users of filtering. Information is filtered based on user profile and additive user profile for experiential users of filtering. The overall research effort has been broken down to a set of detailed research works:User profile is created based on ontology. User profile is expressed by topics, which is defined by ontology. Information share about user profile is realized. Based on user profile, topic template and additive user profile are created.Topic template, which expresses users' common information demand, is brought forward. In users of filtering, some no experiential users cannot exactly express their demand, cannot get good results of filtering. Based on user profile, topic template, which is created, assists user profile for information filtering. Users will get good results of filtering.Additive user profile, which expresses information demand recommended by users whose demand is alike, is brought forward. In users of filtering, some experiential users can expressly express their demand. But these users usually need to comprehend other's view—other user demand. Based on user profile, additive user profile is created by means of users whose demand is similar to user demand of filtering. Additive user profile recommends information demand for user of filtering, and assists user profile for information filtering. Users will get good results of filtering.Formula is quantitatively given to judge user type. A user whose time is engaged by a topic is bigger than or equal to the average time engaged by the topic is experiential, or else not experiential.An information filtering method, based on user profile and topic template, is brought forward for no experiential users of filtering. In filtering, user profile, topic template are arranged a proper function.An information filtering method, based on user profile and additive user profile, is brought forward for experiential users of filtering. In filtering, user profile, additive user profile are arranged a proper function.