The paper of note, titled Jigsaw: Supporting Investigative Analysis through Interactive Visualization, was co-authored by Stasko, Carsten Görg, Zhicheng Liu, and Kanupriya Singhal.

The team’s 2007 research developed a visual analytic system, called Jigsaw, which addresses a challenge investigative analysts face when working with large collections of text documents: Connecting embedded threads of evidence to formulate hypotheses. As the number of documents and concepts in such cases grows larger, making sense of the information becomes more difficult.

Jigsaw represents documents and their contents visually in order to help analysts examine reports more efficiently and develop theories about potential actions more quickly. The system performs rudimentary text analysis including sentiment detection, similarity comparison, and clustering, among other tasks, on the documents and then provides multiple interactive visualizations of the documents’ text. It provides multiple coordinated views with emphasis on visually illustrating connections between entities across the different documents.

“This was probably the biggest project in my lab, maybe, over my entire career in terms of how many students were on it,” Stasko said. “It was a big effort for maybe seven or eight years, and this paper was our first introduction of the idea.”

Stasko’s lab published a number of subsequent papers relating to the system (a list can be found here).

IEEE VIS is being held Oct. 1-6 in Phoenix, Ariz., and includes a number of co-located conferences and programs, including IEEE VAST, IEEE Information Visualization, and IEEE Scientific Visualization.

A full account of Georgia Tech’s participation at the conference can be found here.