I. Introduction
Social media (e.g., Twitter and Weibo) has provided an effective means for people to share and comment on the events happening around them [1] – [3]. Event detection on social media is a common task in visualization. An event is defined as a 4-tuple 〈who, what, where, when〉, indicating specific people who discuss specific content at a location and time point [4]. Events always receive much attention, with many posts discussing the same content intensively published in a short time frame. The content describes a real-world object such as a sports game, disease, or a natural phenomenon. Thus, event detection commonly relies on continuous monitoring of massive quantities of multivariate data [4].