Sub-Events Tracking from Social Network based on the Relationships between Topics

Online social networks like Twitter have emerged as the most popular way to discuss what is happening about real-world events. The topics related to the event may change over time due to several factors. People’s interests in the event are changed over a period of time. This paper we proposed an approach for sub-events tracking in micro-blog service (i.e., Twitter). Our algorithm can identify sub-events and track a sequence of events in chronological order. We utilizes an approach of Latent Dirichlet Allocation (LDA) for sub-event detection. Sub-event tracking task is proposed based on the relationships between sub-events by adopting community evolution discovery techniques. In our experiments, we demonstrate the applicability of our method by using Twitter messages related to the 2018 Protests in Paris.