Understanding Public Emotional Reactions on Twitter
Keywords:emotion, text analytics, visualisation
Social media has become an important communication medium for people to express their feelings and opinions, sometimes on a daily basis. We demonstrate a system that helps monitor and understand how people feel about specific topics or events. Using a large vocabulary of emotion words constructed from a number of public domain resources and mapping them to an emotion hierarchy, our system is able to analyse incoming posts gathered on a topic of interest and visualise the emotional pulse of a geographical region as it reacts to that event. The content analysis component of the tool then enables users to gain insights into the mood variations exhibited. We illustrate the system by exploring the public reaction to a specific recent event: the Sydney siege tragedy on December 15th, 2014.