Abstract
In this demo paper we present EmoTrend, a web-based system that supports event-centric temporal analytics of the global mood, as expressed in Twitter. Given a time range, and optionally a set of keywords, the system relies on peak frequencies, and the social graph, to identify relevant events. Subsequently, by performing sentiment analysis on related tweets, the global impact and reception of the events are presented by a visualization of the overall mood trend in the time range.