Movie Recommendation System for Educational Purposes Based on Field-Aware Factorization Machine

With rich resources, movies have been applied as instructional media in the domain of education, such as fields of Second/Foreign Language Leaning, Communication, and Media Art. Factorization machine (FM) can effectively simulate common matrix factorization models by changing the form of real-value vector, which can be utilized in movies recommendation under the context of education. However, it is usually used to solve classification tasks. This paper applies the field-aware factorization machine (FFM) to solve movie rating prediction and help users select appropriate movies for learning purposes. In order to further enhance the availability of the model, clustering algorithm is also integrated in FFM for adding new fields. The experimental results demonstrate the effectiveness of the proposed methods in reducing the RMSE.