Abstract
Sparsity is a tough problem in a single domain collaborative filtering (CF) recommendation system as it is difficult to compute the similarities among users accurately. Recently, cross domain CF is a new way to alleviate this difficulty. In this paper, we propose a user-based cross domain CF algorithm based on a linear decomposition model. We pour the items together and learn a linear decomposition model to explore the relationship between the total similarity and the local similarities of different domains. We first construct training samples by computing the similarities of any two users in different domains. Then, we solve a linear least square problem to obtain the decomposition coefficients. Finally, we compute the local similarity in the target domain using the decomposition model. Since we compute the similarity in the target domain with the help of rich ratings in other domains, this similarity would be expected to be more accurate than the measured similarity computed by the sparse ratings in the target domain. We conduct extensive experiments to show that the proposed algorithm is effective in addressing the data sparsity problem, as compared with many state-of-the-art CF methods.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.