The GTD Data Rivers is an interactive visual exploratory tool that allows users to investigate temporal trends in terrorism in the Global Terrorism Database (GTD). The GTD Data Rivers aggregates important variables from the database and visualizes them as a comprehensible stack chart.
Five different stack charts can be selected:
A stack chart analyzes every incident in the GTD, both domestic and international, from 1970 to 2010 (over 98,000 discrete events) and aggregates them according to the selection of the user. A unique layer is created for each data stream, with each stream reflecting a value of the variable being displayed. The thickness of each layer changes along the horizontal axis, representing its frequency in the database.
As an example, the default chart (Countries Attacked), creates a layer for every country in the GTD that has experienced a terrorist attack at some point since 1970. Thick layers represent countries that have had many attacks, whereas thin layers signify countries that had fewer attacks. Users can mouse-over a layer to see the country it represents and quickly see how the layer thickens or thins over time. In this example, users can also use the Search Box to look for specific countries, or filter according to incident counts. One of the benefits of stack charts is that aggregate trends emerge, as layers stack up like a histogram.
Additional information on the development of the GTD Data Rivers is available in the paper "Exploring Global Terrorism Data: A Web-based Visualization of Temporal Data." (J. Lee, “Exploring Global Terrorism Data: A Web-based Visualization of Temporal Data,” ACM Crossroads, 15, 2, 7-16, 2008.)
The GTD Data Rivers was developed as a joint effort between the Human-Computer Interaction Lab and START, both at the University of Maryland. Development efforts were led by Joonghoon Lee (firstname.lastname@example.org), under advisement from Ben Shneiderman and Adam Perer.