Modeling and Analyzing of Research Topic Evolution Associated with Social Networks of Researchers

Research trends keep evolving along the time with certain trackable patterns. Mining academic literature and discovering the latent research trends evolution is an interesting and important problem. Few of previous studies focusing on academic topic evolution modeling have addressed the temporal topic evolution patterns. In addition, researchers' profile and their social networks are valuable complementary to the research trends tracking. In this study, to analyze the underlying research trends evolution along with the scientific collaborations of researchers, a novel temporal research trends evolution model associated with researchers' social networks is proposed and built. Specifically, the detected research topics are classified into different clusters in each timeslot, and the evolution patterns are deduced among these topic clusters. The effectiveness of our approach is evaluated based on a real academic dataset. The experimental results can help users to discover the major research trends for specific fields. Besides, the tracked statuses of the corresponding scientific groups are helpful for searching research trends or finding collaboration opportunities according to researchers' different requirements.