Online Link Strength Trend Model Based on Content and Topology

Previous work analyzing social network has focused on link existence of friendship relation, but not the strength of link for item recommendation and information diffusion. The link strength reflects a person´s interesting which is influenced by neighbors or particular content, so the root cause of users´ preferences needs to be mining for improving the accuracy of personal recommendation. With this aim in mind, we propose a novel Content -Topology Time model (CTT) for link strength predicting in time-aware online social network. CTT model integrates topological features, content semantics and time to monitor pair relationship. Firstly, an online version Author-Recipient-Topic model (OART) is developed to track the content evolution of pair relationship for text flies stream. Moreover, interest source mining is exploited to understand users´ communication behavior in our model. Comprehensive experimental studies on real-world micro-blog data set show that our approach outperforms existing ones and well matches the practice.