Cooperative Recommendation Based on Ontology Construction

In the traditional information service system, the lack of service personalization and semantic information influences the intelligence of information service, and decreases the quality of service. Cooperative recommendation has emerged as an effective technology for personalized information access. In this paper, we present a framework of cooperative recommendation based on ontology construction. We first construct domain ontology, and find user interest through user behavior record database, then construct user interest ontology by the mapping relationship between user interest and domain ontology. Through identifying the vertical weight by user interest similarity, the horizontal weight by the level of ontology concept utilize counted by PageRank, we use the improved weighted association rule algorithm to mining the concepts in user interest ontology, and exploit the user interest model. At last, we realize the semantic-based, personalized cooperative recommendation oriented to content.