Browse Prior Art Database

Interactive Visual Synthesis of Analytic Knowledge

IP.com Disclosure Number: IPCOM000187744D
Original Publication Date: 2009-Sep-17
Included in the Prior Art Database: 2009-Sep-17

Publishing Venue

IBM

Abstract

We have created a novel system supporting interactive, progressive synthesis of analytic knowledge. Unlike existing visual analytic tools, which typically support only exploration of existing information, our system offers two unique features. First, we support user-system cooperative visual synthesis of analytic knowledge from existing data. Second, we support progressive visual analysis of synthesized knowledge.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 9% of the total text.

Page 1 of 14

Interactive Visual Synthesis of Analytic Knowledge

A visual investigation involves both the examination of existing information and the synthesis of new analytic knowledge. This is a progressive process in which newly synthesized knowledge becomes the foundation for future discovery. We have created a novel system supporting interactive, progressive synthesis of analytic knowledge. Here we use the term ``analytic knowledge'' to refer to concepts that a user derives from existing data along with the evidence supporting such concepts. Unlike existing visual analytic tools, which typically support only exploration of existing information, our system offers two unique features. First, we support user-system cooperative visual synthesis of analytic knowledge from existing data. Specifically, users can visually define new concepts by annotating existing information, and refine partially formed concepts by linking additional evidence or manipulating related concepts. In response to user actions, our system can automatically manage the evolving corpus of synthesized knowledge and its corresponding evidence. Second, we support progressive visual analysis of synthesized knowledge. This feature allows analysts to visually explore both existing knowledge and synthesized knowledge, dynamically incorporating earlier analytic conclusions into the ensuing discovery process.

INTRODUCTION

Information analysis is a complex task in which analysts must make
sense out of a sea of heterogeneous, potentially relevant information.
In such a process, an analyst must examine existing information
as well as develop analytic knowledge to progressively achieve
their investigational goals. For example, a law enforcement analyst
investigating an alleged case of political corruption would need
to examine existing data, such as telephone records and newspaper
articles, and use such information to synthesize new knowledge
about potential suspects, including their social networks and political
motives. Using their synthesized knowledge, the analyst can
then begin to uncover the overall conspiracy.

Along with information exploration, therefore, analysts
must also perform knowledge synthesis to gradually build an
understanding of concepts or events that are only indirectly supported
by the raw information. In this process, analysts must have
tools for defining newly derived knowledge, as well as tools for
modifying previously synthesized knowledge in response to new
discoveries. As shown in Figure 1(a), the analyst investigating the
alleged political fraud uses our tool to mark the bio-tech startup
called Boynton as a potential scam, and associate it with supporting
evidence, such as the FDA investigation event indicated on the
timeline.

Furthermore, information analysis is a progressive process in
which data exploration and knowledge synthesis occur iteratively
as an i...