SOS: A multimedia recommender System for Online Social networks

The use of Online Social Networks has been rapidly increased over the last years. In particular, Social Media Networks allow people to communicate, share, comment and observe different types of multimedia content. This phenomenon produces a huge amount of data showing Big Data features, mainly due to their high change rate, large volume and intrinsic heterogeneity. In this perspective, in the last decade Recommender Systems have been introduced to support the browsing of such data collections, assisting users to find "what they really need" within this ocean of information. In this research work, we propose and describe a novel recommending system for big data applications able to provide recommendations on the base of the interactions among users and the generated multimedia contents in one or more social media networks. The proposed system relies on a "user-centered" approach. An experimental campaign, using data coming from many social media networks, has been performed in order to assess the proposed approach also showing how it can obtain very promising results.