A Random Forest Regression-based Personalized Recommendation Method
Yingxue Ma, Mingxin Gan · EasyChair preprint · 2018
Recently, recommendation methods have achieved remarkable success are based on the similarity between users or objects. However, high similarity between users does not represent they have similar preference in reality. What really reflects user preferences is the user's subjective rating on the item. In this paper, we propose a method to predict users' ratings of films by using random forest regression. As the user's rating on items depends on the characteristics of the item and the preferences reflected in history records, we use these two data as input, the user's scoring process is simulated under the random principle to predict the user's rating on the film. The results show that our proposed method outperforms others in MAE.