Research On Cold-start Problems In Collaborative Filtering Systems

Since the mid 1990s, the first thesis on Collaborative Filtering algorithm appears, recommender systems become an important area of research. Collaborative Filtering recommender systems successfully solve the problem of the network information overload, users can find what their needs to take a short time. Although Collaborative Filtering (CF)-based recommender systems have received great success in a variety of applications, they still under-perform and are unable to provide accurate recommendations when users and items have few ratings, resulting in reduced coverage.To overcome these limitations, we propose an effective hybrid user-item trust-based recommendation approach in this paper that fuses the users'and items' implicit trust information. We have also considered and computed user and item global reputations into this approach. This approach allows the recommender system to make an increased number of accurate predictions, especially in circumstances where users and items have few ratings. Experiments on one real-world dataset show that the proposed recommendation approach significantly outperforms state-of-the-art recommendation algorithms in terms of recommendation accuracy and coverage, as well as significantly alleviating data sparsity, cold-start user and cold-start item problems.