Making Sense of the Immense: An Application of Big Data Visual Analytics for Spatiotemporal Sensemaking in Social Media
This work describes SensePlace2, a web-accessible geovisual analytics tool that allows analysts to achieve greater situational awareness about key topics of interest. SensePlace2 collects, analyzes, and visualizes an immense and growing number of tweets that analysts can explore through ad hoc queries and a rich set of user interactions. We pair entity extraction and geocoding algorithms to determine the context and topic of tweets, the locations being referred to by the tweets, and how these topics and relationships among them vary across time and space. SensePlace2 is designed with multiple coordinated views so that users can query, filter, select, and manipulate tweets across both geographic and temporal domains. During an analysis session queries are stored in a history that can be returned to at any time. The system also employs feedback mechanisms that allow analysts to report errors directly to the development team. Work is currently underway to enrich the context foraging capabilities of SensePlace2 by connecting the existing database to ancillary data sources and information.