A Survey on the Scalability of Recommender Systems for Social Networks

Abstract

It is typical among online social networks’ users to share their status, activity and other informationwith fellowusers, to interact on the information shared by others and to express their trust or interest for each other. The result is a rich information repository which can be used to improve the user experience and increase their engagement if handled properly. In order to create a personalized user experience in social networks, we need data management solutions that scale well on the huge amounts of information generated on a daily basis. The social information of an online social network can be useful both for improving content personalization but also for allowing existing algorithms to scale to huge datasets. All current real-world largescale recommender systems have invested on scalable distributed database systems for data storage and parallel and distributed algorithms for finding recommendations. This chapter, focuses on collaborative filtering algorithms for recommender systems, briefly explains how they work and what their limitations are.