Pattern-Based Semantic and Temporal Exploration of Social Media Messages

Social media is a valuable source for emergency workers and first-responders due to its wide use. Unfortunately, the number of posts poses a challenge to quickly find relevant and trustworthy information. We propose a text pattern-based approach to gain insights from large micro-document collections, including temporal and semantic relationships. In this work, we apply our method to the fictional data set of the VAST 2019 Mini-Challenge 3 that deals with the aftermath of an earthquake. We present major findings we could deduct using our visual analytics approach.