Publication

Abstract : Automated systems like Recommenders are developed to recommend item suggestions to consumers abiding a variety of conditions. They predict ratings of the most likely product a user will buy intuitively or based on interest. Collaborative filtering (CF) is one of the most common methods for designing recommender systems. It works on a collection of users’ established preferences to make suggestions or predictions for unknown users. This survey study provides a comprehensive insight into the most up-to-date state-of-the-art approaches for efficient filtering and successful recommendation. In addition, the survey article studies different types of deep learning-based collaborative filtering algorithms and implementation techniques used in different recommender system applications to overcome cold start and data sparsity challenges.