Enhanced Deep Autoencoder based Recommender System
Karima Mecheri, Rostom Mamadji, Sihem Klai, Labiba Souici‐Meslati · 2022
With the colossal volume of information on the Internet, it is becoming more and more difficult for users to find the desired information quickly and easily. To surmount this problem, recommender systems have been developed as an effective solution. However, the main challenges faced by these systems are data scarcity and their constantly increasing dimensions. Recently, several researchers have used deep learning to solve these problems. In this paper, we first propose a movie recommendation system based on collaborative filtering using the auto-encoder. The role of this approach is to overcome the data scarcity problem and therefore improve accuracy. To enhance our approach we use pre-trained RBM (Restricted Boltzmann Machine) models to build our second Deep Auto-Encoder. To validate our approach, we conduct experiments on a set of real data, and we compare the two systems using evaluation metrics.