Collective Sensemaking via Social Sensors: Extracting, Profiling, Analyzing, and Predicting Real-world Events

Social media platforms like Twitter and Facebook have emerged as some of the most important platforms for people to discover, report, share, and communicate with others about various public events, be they of global or local interest (some high profile examples include the U.S Presidential debates, the Boston bombings, the hurricane Sandy, etc). The burst of social media reaction can be seen as a valuable real-time reflection of events as they happen, and can be used for a variety of applications such as computational journalism. Until now, such analysis has been mostly done manually or through primitive tools. Scalable and automated approaches are needed given the massive amounts of both event and reaction information. These approaches must also be able to conduct in-depth analysis of complex interactions between an event and its audience. Supporting such automation and examination however poses several computational challenges. In recent years, research communities have witnessed a growing interest in tackling these challenges. Furthermore, much recent research has begun to focus on solving more complex event analytics tasks such as post-event effect quantification and event progress prediction. This tutorial aims to review and examine current state of the research progress on this emerging topic.