The explosive growth of social network with the help of e-commerce websites has made the issue of information retrieval increasingly challenging. Users may not have the time or knowledge to personally evaluate the options which are available on social network platform. Recommender systems present themselves as a practical answer to endless options available online. In this research paper, we give an insight into various types of filtering techniques associated with recommender systems. We also discuss the problems faced by these filtering techniques and evaluation criteria on the basis of which various algorithms made for recommendation purpose can be compared. Finally we propose a composite news recommendation system model. © 2015 IEEE.