Discovering Communities of Users on Social Networks Based on Topic Model Combined with Kohonen Network

Interaction among users on social networks through messages and interested topics forms online communities. The question is how to discover what communities users belong to or what online communities are interested in or what each period of time the interested topic change in online communities are? To answer these questions, this paper proposes a new model for discovering communities on social networks based on the topic model combined with Kohonen networks. This model, we focus on discovering online communities and surveying the changes in interested topic and users in communities with temporal factor. The proposed model is experimented with a set of interested topic vectors. These topics are exploited from a corpus of messages in Vietnamese on social networks in the higher education field.