RSAttAE: An Information-Aware Attention-Based Autoencoder Recommender System

Recommender systems play a crucial role in modern life, including information retrieval, pharmaceutical industry, retail, and entertainment. The entertainment sector, in particular, attracts significant attention and generates substantial profits. This work proposes a new method for predicting unknown user-movie ratings to enhance customer satisfaction by recommending most relevant products to users. We propose RSAttAE, an attention based autoencoder architecture which leverages attention mechanism to incorporate entities' features enabling learning more meaningful and robust representations for users and movies from rating matrices. Furthermore, XGBoost regressor was utilized for rating predictions using learned representations. Our experiments on MovieLens 100 K demonstrate that our proosed method outperforms most of the existing state-of-theart methods. The code is available at github.com/ComputationIASBS/RecommSys.