Loading...

Combining Trust-Based and Collaborative Filtering Methods to Enhance Recommender Systems

Foroughi Dehnavai, Sobhan | 2015

1501 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 47623 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Beigi, Hamid
  7. Abstract:
  8. Nowadays, recommender systems have become powerful tools that engage users in an online manner, over the Internet. Collaborative filtering (CF) is a well established method for building recommender systems and has been applied to several applications. While CF has its advantages,its use is hindered by challenges such as low accuracy for new users (newcomers). With the growth of online social networks, networkbased recommender systems emerged. These systems take advantage of the information available in social networks and the user’s past activity to recognize user behavior and recommend items that are more relevant to each user. One of the most important advantages of network-based recommender systems is their higher accuracy for newcomers, due to the fact that information extracted from the user’s social interactions allow the system to identify similar users and hence recommend items that are more likely to interest the user.In this work, a novel model-based approach is presented to improve recommender systems by using information available from online networks. In this method, each item/user is modeled using a vector in a latent subspace, such that a user’s interest in an item is equivalent to the inner product of the respective vectors. To introduce the information extracted from social networks into the proposed model,weighted sparse representation is used to model user vectors.The parameters of this model are determined by maximum a posteriori estimation.To evaluate the method empirically, its experimental results are compared with state-of-the-art network-based recommender systems over Epinions and Flixter databases. The results show that the proposed method provides superior performance in comparison
  9. Keywords:
  10. Trust Network ; Matrix Factorization ; Collaborative Filtering ; Recommender System ; Weighted Sparse Representation

 Digital Object List

 Bookmark

...see more