The Intelligent Recommendation System Based on Amended Rating Matrix in TTP

As an important technology transfer media, technology transfer platform (TTP) is being paid extensive attention. Collaborative filtering recommendation method is one of the important recommendation strategies, and is also the most successful one ever applied till now. Traditional collaborative filtering methods, unavoidably, are puzzled by the problem of "data-sparsity"; and in this paper, the intelligent recommendation system is brought forward to settle the problem. The system firstly acquires user's implicit degree-of-interest through analysing Web log; and then combines the implicit degree-of-interest and the explicit rating matrix. On the basis of mixed rating matrix, semantic classifying information is imported to decrease the rating matrix's dimensions and thus further improve the precision of collaborative recommendation. After collaborative recommendation amended, we propose a novel architecture of intelligent recommendation system in TTP. The system has achieved expected effect as it is shown in an experiment