Heterogeneous Ensemble based Collaborative Filtering (HECF) Model for Movie Recommendation
Recommendation systems have become major driving factors of user decision making. Hence organizations tend to incorporate them into their business models to reduce the design fatigue and to provide customer satisfaction. This paper presents a heterogeneous ensemble based collaborative filtering (HECF) model for movie recommendations. The proposed model is composed of multiple prediction models creating a heterogeneous ensemble. User requirements are analyzed and similar user requirements are also incorporated to create the training data. Predictions from the model are then integrated to form a single unified prediction. Experiments in terms of MAR and RMSE and comparisons with existing models exhibit the high efficiency of the proposed HECF architecture.
Index Terms - Collaborative Filtering, Movie Recommendation, Ensemble, User Similarity, Movie Rating