Customizing knowledge-based recommender system by tracking analysis of user behavior

In this paper, we reviewed the major problems in the existing recommender systems and presented a tracking recommender approach based on user´s behavior information and two-level property of items. Our proposed approach defined user profile model, knowledge resources model and constructed Formal Concept Analysis (FCA) mapping to guide a personalized recommendation for user. We simulated a prototype recommender system that can make the quality recommendation by tracking user´s behavior. The experimental result showed our strategy was more robust against the drawbacks and preponderant than conventional recommender systems.