Recommender System based on Deep Neural Network and Long Short Term Memory

To provide relevant recommendations for clients, a recommendation system is essential in online commerce, streaming services, and news article websites. Existing methods in recommendation systems are limited by the cold start problem. The Deep Neural Network (DNN) – Long Short-Term Memory (LSTM) technique is developed in this study to improve the efficiency of recommendation systems. The DNN method is used to predict new user ratings based on prior user ratings, while the LSTM method is used to recommend a relevant movie to the user. The user-item similarity was calculated and used in the LSTM algorithm to offer the relevant recommendation. The LSTM approach has the advantage of storing relevant information over time and making appropriate recommendations. The proposed DNN-LSTM (Deep Neural Network-Long Short-Term Memory) technique in the recommendation system is evaluated using the MovieLens 100k and 1M datasets. In the MovieLens 100k dataset, the proposed DNN-LSTM approach has an RMSE of 0.431, while the existing HCBCF (Hellinger Coefficient Based Collaborative Filtering) method has an RMSE of 0.871.