Investigative analysts and
researchers acquire clues and connect small bits of evidence to
uncover larger plans, stories, or narratives, and to simply gain a
better understanding of the information. Often, the individual bits
of evidence are short text documents or spreadsheets, and analysts
must examine large collections of such documents in order to "put the
pieces together" and formulate a well-supported hypothesis about
actions that may occur in the future. As the number of documents to
examine rises, it becomes more and more challenging for analysts to
understand the data and make judgments about it.
We have created Jigsaw, a visual analytics system to help analysts and
researchers better explore, analyze, and make sense of such document
collections. Our specific objective is to help analysts reach more
timely and accurate understandings of the larger stories and important
concepts embedded throughout textual reports. Jigsaw provides a
collection of visualizations that each portray different aspects of
the documents. We particularly focus on presenting the identifiable
important entities (people, places, organizations, etc.) and their
direct or indirect connections. Textual processing extracts the
important entities from the documents and then the visualizations help
an analyst to explore the relationships and connections among the
entities. The system includes a variety of visualizations such as
list, graph, temproal and connection-based views, as well as views of
individual document's text and the document collection as a whole.
Jigsaw essentially acts as a visual index onto the document
collection, helping analysts identify particular documents to read and
We have used Jigsaw to explore a wide variety of domains and document
collections including academic papers, grants, product reviews,
business press releases, news articles, intelligence and police
reports, statutes, and even books such as the Bible. Jigsaw is
available for you to try and
use on your own documents and data.
For a good, short (2-page) overview of Jigsaw's goals and methods,
please see the Spring 2008 VacViews
article about the system. We have written and presented a number
of academic papers about Jigsaw. To gain a good understanding of
Jigsaw, please read the
introductory 2008 Information
Visualization journal article and the more
recent 2013 IEEE
TVCG journal article about how we combine computational
analysis of document text with interactive visualization in the
system. To learn more about the system's details and its different
visualizations, please examine the Jigsaw views
page. And to see the system in action, explore the scenario
videos listed up above on this webpage or examine
the video tutorial page.
Using Jigsaw as an analytic aid, we entered and won the university
division of the VAST 2007
Contest. To learn more about our contest entry, visit the Contest Summary page.
The picture below shows Jigsaw being used on a computer with four monitors
presenting the different system views. The multitude of views in the
system makes a mutliple monitor computer like this desirable for analysis.
If four monitors aren't enough, Jigsaw can really hum on a display
wall. The picture below shows Carsten demonstrating Jigsaw running on
the PowerWall at Konstanz University (Germany) in Daniel Keim's lab.
Below is a picture of John demonstrating Jigsaw to Tim Collins of
Purdue at the 2009 NVAC Consortium meeting at PNNL.
This research is supported in part by the VACCINE DHS Center of
Excellence and NSF Awards IIS-0915788 and CCF-0808863 (FODAVA Lead).
Past support for the project came from a grant from
the Dept. of Homeland Security's NVAC
Program and NSF Award IIS-0414667.
modified: January 20, 2014