A Comparative Study on Recommender Systems Approaches
Kawtar Najmani, El Habib Benlahmar, Nawal Sael, Ahmed Zellou · 2019
Recommender systems aim to generate interest items or products for web users. They offer useful information and adapted to users' profile based on their preferences and behaviors. Currently, they are proposed in several domains, namely: e-commerce, e-learning, research, music, social networks, etc. There are mainly three approaches that are used in the recommender systems, those based on content, those based on collaborative filtering, and finally the hybrid approaches, which merge different algorithms and provide more accurate and effective recommendations than a single algorithm, as the disadvantages of one simple technique can be overcome by another technique. This paper offers a comparative study of approaches in recommender systems, starting with a general presentation of each one, then it treats the advantages, the limitations, and the techniques, which give a high accuracy in each approach.