Event-based trend factor analysis based on hashtag correlation and temporal information mining

© 2018 Elsevier B.V. Nowadays, using social media such as Twitter, Facebook, etc., has become extremely popular among individuals around the world. Utilizing various event analysis algorithms for research on which event/events is/are the hottest as well as discovering the reason(s) behind it/them. Based on our proposed model, we can investigate about all of the reasons of the events and why they triggered the event(s) to a comprehensive discussions. In addition, we can list the reason's impaction from the highest one to the lowest one. The idea of event-based analysis is that we can access a good explanation on social interactions and behaviors associated with complex situations. Previous studies can be simply categorized into two parts. One part is the discovery of event clusters with different temporal concerns, and the other part is the collection of related event(s) and the calculation of correlated connection strength among them. But both parts are only focused on the lack of notice the reason why these events have been discussed. Our proposed model is searching for the deeper issue, not only the idea behind the event, but also the reason why it makes the event triggered the comprehensive discussion. Furthermore, different from the clustering algorithm, our search layer can be increased as many as we need until reaching the goal. We demonstrate our model for constructing a strength table that contains the reasons related to the event, and the result can be presented precisely either as a table or a graph for user's easy-understanding.