Social Media Recommendation

Social media recommendation is foreseen to be one of the most important services to recommend personalized contents to users in online social network. It imposes great challenge due to the dynamical behavior of users and the large-scale volumes of contents generated by the users. In this chapter, we first present the principal concept of social media recommendation. Then we present the framework of social media recommendation, with a focus on two important types of recommendations: interest-oriented social media recommendation and influence-oriented social media recommendation. For each case, we present the design of the recommendation that takes both social property and content property into account, such as user relations, content similarities, and propagation patterns. Furthermore, we present theoretical results and observations on the social media recommendation approaches. 1 Social Media Recommendation in Online Social Network Online social network is attracting more and more people in today’s Internet, where users can share and consume all kinds of multimedia contents. Social media sharing is based on online social network, where users are able to reach various contents shared by others. With the exponential growth in social media contents, such as images and videos generated by users, it is of great importance to study how to provide personalized contents in the social media service. Recommendation is foreseen to be one of the most important services that can provide such personalized multimedia contents to users. However, social media recommendation is different from tradiZ. Wang W. Zhu, P. Cui, L. Sun and S. Yang are with the Department of Computer Science and Technology, Tsinghua University, Beijing, China, 100084, e-mail: {wangzhi04@mails., wwzhu@, cuip@, sunlf@, yangshq@}tsinghua.edu.cn