Classification and applications of social recommender systems

Abstract

The rapid growth of internet has paved the way for recommender systems as users are inundated with a variety of choices. A computer system that provides recommendations/suggestions to the users is called Recommender Systems (RS). These systems are a subclass of information filtering systems that provides suitable recommendations to users about a product, movie, song, mobile applications etc. RS are being widely used in social networking sites like Facebook, Twitter, LinkedIn etc. to help users choose product/friend/job from the various recommendations. These systems analyze patterns of user interest in posting information, sharing posts or joining a new group to provide personalized recommendations that suit user’s taste. RS are popular in many areas especially for predicting/suggesting what the users want in advance and thus help in reducing search costs. Recommendations are done using either collaborative, Content-based filtering or hybrid approach. Content-based recommendation system recommends based on user profile created by analyzing the items the user has liked/rated in the past. A collaborative recommendation system, on the other hand, does suggestions based on the preferences/ratings of our previous collaborators (users whose liking was similar to ours). The popularity of social networking sites is increasing with more number of users being added daily. This paper discusses the various classification and applications of RS in some of the popular networking sites.