A Knowledge-based Recommendation Framework using SVN Numbers

Current knowledge based recommender systems, despite proven useful and having a high impact, persist with some shortcomings. Among its limitations are the lack of more flexible models and the inclusion of indeterminacy of the factors involved for computing a global similarity. In this paper, a new knowledge based recommendation models based SVN number is presented. It includes database construction, client profiling, products filtering andgeneration of recommendation. Its implementation makes possible to improve reliability and include indeterminacyin product and user profile. An illustrative example is shown to demonstrate the model applicability.