Recommender Systems: Algorithms, Evaluation and Limitations

Aims/ objectives: This paper presents the different types of recommender filtering techniques. The main objective of the study is to provide a review of classical methods used in recommender systems such as collaborative filtering, content-based filtering and hybrid filtering, highlighting the main advantages and limitations. This paper also discusses the state-of-art machine learning based recommendation models including Clustering models and Bayesian Classifiers. Further, we discuss the widespread application of recommender systems to a variety of areas such as e-learning and e-news. Finally, the paper evaluates the performance of matrix factorization-based models, nearest neighbours algorithms and co-clustering algorithms in terms of different metrics.