Abstract

With the growth of the media market, companies that provide contents services such as movies, music and video are providing various content to satisfy users. While these changes have allowed users to enjoy richer content, a new problem has emerged that they have to spend much more time than before to find content that suits their taste among the overflowing content. Recommender system has become an important key to solve these problems. Matrix Factorization (MF) is the most well-known and widely used for identifying users’ preference on contents. However, MF has a drawback of data sparsity and is not capable of utilizing meta-data. In this study, we proposed a two-stage contents preference model with Matrix Factorization (CPMF). The proposed method combines MF and contents preference models that utilize a variety of meta-data (e.g., actors, directors, and genres) to identify users’ preferences. A numerical analysis is conducted to evaluate the performance of the proposed method for movie recommendation domains.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call