Understanding research topic evolution in a specific field is important to learn the temporal structure and latent research trends. Although research works in terms of topic analysis have been developed for many years, it is still difficult to well understand the topic evolution due to the limitation of the bibliographical data which does not consider the researchers' profiles and their social networks. In this study, to analyze the underlying research topic evolution and facilitate the scientific collaborations for researchers, a novel topic evolution model based on the analysis of researchers' social networks is built. Specifically, the researchers' co-authorships associated with their latent social networks are analyzed. A research topic tracking method is then developed to track the topic evolutions and analyze the research lifecycle for the latent scientific teams. The derived results can help users to discover the major research trends for specific fields. And the tracked statuses of the corresponding scientific groups are illustrated in different lifecycle stages, which is useful for the search of research trends or finding of collaboration opportunity according to researchers' different requirements.