Intelligent User Interfaces for Intelligence Analysis

Background and Motivation

Intelligence analysis is a process of evaluating and transforming raw data into descriptions, hypotheses, and explanations for intelligence consumers. For example, a government agency may gather intelligence about an organization or individuals to investigate a possible terrorist attack, or a financial agency may gather business and market intelligence to evaluate a potential investment candidate. An intelligence analyst is an individual who performs intelligence analysis of a given situation. Using the above examples, a counter narcotics agent or a financial analyst is an intelligence analyst.

Intelligence analysis is a complex activity due to the nature of the task, inherent limitations in human cognitive processes [1], and the environment. First, intelligence analysts must deal with much uncertainty in the process due to often incomplete and inconsistent information. Second, human analysts suffer from reasoning and judgment failures as well as cognitive biases to include anchoring, over-generalization, and so on. The difficulty of an analysis task increases when analysts must deal with large amounts of dynamically updated, transient information (e.g., surveillance videos) in very short time frames. Finally, the environment may be one in which access to intelligence sources is denied and/or active deception techniques are applied against the analyst?s discovery of the truth.

To tackle such a difficult task, in practice analysts normally take a target-centric approach [2]. Given an analysis task/objective, analysts come up with multiple hypotheses and validate the hypotheses by gathering and distilling information from various sources. A hypothesis is a tentative assumption made by an analyst in order to draw out and test its logical and empirical consequences. For example, a counter terrorism agent may speculate on the location or participants of a possible terrorist attack or a financial analyst may assume the demand and supply trend of an industry. During a hypothesis validation process, relevant information would be retrieved as evidence that either supports or refutes a hypothesis.

As intelligence analysis becomes an increasingly important area for many sectors (e.g., government and financial), researchers and practitioners have developed a wide range of computer technologies and tools to aid human analysts in performing their tasks. However, most of the efforts focus on information search (e.g., QA-based information retrieval) or information analytics (e.g., data mining and video analysis). Limited attention has been devoted to providing better user interfaces through which human analysts can better manage a complex analysis process. With the current far-from-perfect information search and analytic technologies, we believe that it is critical to create a ?smart? interactive environment where human analysts and computers can work collaboratively. On the one hand, computers use the human analysis process as a context to better understand the needs of analysts and assist them in their tasks (e.g., reminding the analysts to investigate competitive hypotheses). On the other hand, human analysts could use computer-presented information as a context to better articulate their needs (e.g., issuing follow-up data inquiries) and manage their analysis process (e.g., better organizing hypotheses and evidences). The purpose of this workshop is to bring together researchers and practitioners who are interested in developing and applying the state-of-art intelligent user interface (IUI) technologies to enhance intelligence analysis. Ideally, the developed IUI technologies or tools can capitalize on human strength (e.g., reasoning and detecting visual patterns) and compensate for human weakness in a process of intelligence analysis.

Objectives

Intelligence analysis is a hot topic in multiple sectors including many government funding agencies. Using this workshop as a starting point, we aim to foster closer relationships between a research community and government agencies/companies that are concerned with intelligence analysis. For example, we can identify relevant unclassified data sets for evaluation, share best research practices, propose to establish common research challenge problems, joint programs/consortiums to help fund relevant research efforts, activities to create common standards, or to jointly identify common requirements.

This workshop aims to achieve two main objectives: 1). Identify key research issues and challenges in designing and developing intelligent user interface technologies for intelligence analysis. 2). Bring together researchers and practitioners that are interested in this application area and initiate proper collaboration among different teams.

Organizers

• Dr. Mark Maybury
  http://www.mitre.org/about/technical_centers/itc/about/staff/bio_maybury.html
  Dr. Mark Maybury is executive director of MITRE's Information Technology Division. Mark also serves as Executive Director of ARDA's Northeast Regional Research Center. Mark has over sixty refereed publications. He is editor of Intelligent Multimedia Interfaces (AAAI/MIT Press 1993), Intelligent Multimedia Information Retrieval (AAAI/ MIT Press 1997), New Directions in Question Answering (AAAI/ MIT Press 2004), co-editor of Readings on Intelligent User Interfaces (Morgan Kaufmann Press 1998), Advances in Text Summarization (MIT Press 1999), Advances in Knowledge Management (MIT Press 2001) and Personalized Digital Television (Kluwer Academic, 2004), and co-author of Information Storage and Retrieval (Kluwer Academic 2000).

  Maybury was program chair of ACM's 1999 International Conference on Intelligent User Interfaces (IUI) and co-chair of the 2005 International Conference on Intelligence Analysis. Maybury is a member of the IUI Steering Council, a member of the Board of Directors of the Object Management Group, and Secretary/Treasurer of ACM SIGART. He serves on several international conference program committees and journal editorial boards. He received his B.A. in mathematics from the College of the Holy Cross, an M. Phil. In computer speech and language processing from Cambridge University, England, an M.B.A. from Rensselaer Polytechnic Institute, and a Ph.D. in artificial intelligence from Cambridge University.

• Dr. Michelle X. Zhou
  http://www.research.ibm.com/RIA/People/Zhou/Zhou.htm
  Dr. Michelle Zhou is a research manager at IBM T. J. Watson Research Center, where she manages the department of intelligent multimedia interaction. Her research interests include 3D multimedia user interfaces, intelligent multimodal conversation systems, context-based multimedia presentation authoring, and information visualization.

  Zhou received a Ph.D. in Computer Science from Columbia University in 1999. Her work on automated generation of interactive visual/multimedia explanations has been published in a wide variety of conferences and journals, including CHI, IJCAI/AAAI,ACM Multimedia, and IEEE Visualization/InfoVis. Her recent work on automated generation of information graphics by examples has been featured both in IEEE Computer Magazine, and in the visualization frontier special issue at New York Academy of Sciences. Zhou is also active in research communities both inside and outside of IBM. She was the chair of the multimedia professional impact community (MMPIC) at IBM. She co-organized the international conference on Smart Graphics, and served as program committee members for top-rated conferences, including IEEE Visualization/InfoVis and ACM Multimedia. Her recent work (collaborated with her colleagues) on using an optimization-based approach to address automated multimedia presentation generation won the best paper award at IUI 2005.

Program Committee Members

• Aaron Bobick, Georgia Tech
• Stuart Card, XEROX PARC
• Andrew Cowell, Pacific Northwest National Lab
• Stephen Eick, SSS Research Inc.
• Abigail Gertner, MITRE
• Hari Sundaram, Arizona State University
• Rohan Gunaratna, Institute of Defense and Strategic Studies, Singapore
• David Thurman, Pacific Northwest National Lab
• Zhen Wen, IBM T. J. Watson
• Anita Woolley, Harvard University
• Lisa Yanguas, NSA

Submission Instructions

Submissions may be "regular" papers up to 8 pages long, "position" papers up to 4 pages long, or "demonstration" papers of up to 2 pages illustrating new technology in action. The submission should use the same formatting instructions as the main IUI conference (i.e. ACM style); you can find links to formatting templates at http://www.iuiconf.org/instructions.htm (do not use the automatic submission system at that address).

Mail submissions to mzhou@us.ibm.com. Submissions should include:

1. A separate plain text cover page with title, authors and affiliation, abstract, a list of keywords, and an indication as to whether paper is a regular paper, position paper, or demonstration paper.
2. A PDF file of the paper (regular, position, or demonstration).

Important dates

Contact

Mark T. Maybury
MITRE Corp.
202 Burlington Road
Bedford, MA 01730
maybury[at]mitre.org

References

[1] R. Heuer, Psychology of Intelligence Analysis.
[2] R. Clark, Intelligence Analysis: A target-centric approach.