8 May 2005. Thanks to S.
Sponsored by the Office of the Assistant Director of Central Intelligence for Analysis and Production, this international conference aims to report fundamental, novel, and operationally significant advances in the art of intelligence analysis and the associated science and technology support from national, military, academic and competitive intelligence communities. Analytic communities are continually challenged by the need to analyze massive volumes, velocities, and varieties of multilingual and multimedia data. This situation occurs in multiple disciplines including HUMINT, SIGINT, IMINT, MASINT, OSINT, and GEOINT. This occurs in multiple domains including but not limited to terrorism, politics, economics, chemical, nuclear, and biological weapons of mass destruction, information assurance, science and technology, and industry analysis. Challenges have fueled opportunities for analytic tool developers, educators, and business process owners that support analytic communities in the management of knowledge, information and data sources. Substantial government, commercial and academic R&D activities require a forum for knowledge sharing about intelligence analysis theories, methodologies and tools.
Monday 2 May- Tutorials
Morning Tutorial Session 8 am - 12 noon
(Continental Breakfast 7 am and Coffee Break at 9:45am)
Tutorial 1: Warning Analysis for the Information Age: Rethinking the Intelligence Process (John Bodnar, SAIC)
Tutorial 2: Tailoring Scenario Analysis for Short-Term Policy Challenges (Bill Fiora and Ken Sawka, Outward Insights)
Tutorial 3: Probability and Information: The Essential Facts and Concepts (Dr. Richard Rohwer, Fair Isaac Corp.)
Tutorial 4: Complexity and Uncertainty Management for Better Decision-Making (Gene Allen and Jacek Marczyk, Complex Systems Engineering, Inc)
Tutorial 5: Bayesian Networks: An Analytic Tool for Predicting Behavior (Dennis Buede, Mary Lewellyn, Richard Rees, and Paul Sticha)
Tutorial 6: Intelligent Information Access (Dr. Mark Maybury, MITRE)
Afternoon Tutorial Session 1 pm - 5 pm
(Lunch 12-1 and Coffee Break at 2:45pm)
Tutorial 7: Integrating Methods and Tools to Counter Denial and Deception (Ed Waltz, BAE Systems)
Tutorial 8: The TRA (Threat and Risk Analysis) and the Intelligence Analyst's Contribution (Henry Doucette, Canadian Military College)
Tutorial 9: Role Playing Strategy Games for Intelligence Analysis (Prof. Barry G. Silverman, University of Pennsylvaniaia) - [Max 3 round tables of 10 people]
Tutorial 10: Designing with the Mind in Mind: Basic Phenomena in Human Memory and Problem Solving (Prof. Thomas T. Hewett, Drexel University) [no notes should be published electronically]
Tutorial 11: Event and Temporal Awareness for Intelligence Analysis (Prof.James Pustejovsky, Brandeis University and Inderjeet Mani, Georgetown University)
Tutorial 12: Advanced data mining, link discovery and visual correlation for data and image analysis (Prof. Boris Kovalerchuk, Central Washington University)
Workshop 1: Developing Research Interests of the Homeland Security Center of Excellence for Behavioral and Social Research on Terrorism and Counter-Terrorism. (Prof. Clark McCauley, University of Pennsylvania). Half Day,
Morning 8 am - 12 noon (Continental Breakfast 7 am and Coffee Break at 10:30 am)
Workshop 2: Complex Systems Insights and their Application. Prof. Yaneer Bar-Yam, New England Complex Systems Institute. Afternoon Half day, Afternoon 1 pm - 5 pm (Lunch 12-1 and Coffee Break at 3:30 am).
5:30 - 7:30 Evening: No-host Reception/Cash Bar and Poster-Demonstration Session I (First Floor Foyer - by fountain one level down from lobby)
T1A means Tuesday first session (at 11 am), parallel session A.
W2D means Wednesday second session (2 pm) parallel session D.
W3B means Wednesday third session (4 pm) parallel session B.
Panel 1 - Topic- Culture: The Many Roles of "Culture" in Intelligence: A Panel in Honor of Adda Bozeman.
Chair- Kelcy Alwein, DIA; David Moore, NSA -
Panelists - Kevin O'Connell, DGI; Dr. Pauletta Otis, JMIC; Major Ben Connable, USMC; Dr. Montgomery McFate, ONR; Steve Marrin, UVa; Elizabeth Moore, DoD; Alexandra Hamlet, Aptima.
Panel 2 - Topic- DHS: The Research and Development Agenda for Visual Analytics
Chair - Kristin Cook, Pacific Northwest National Laboratory
Panelists - Jim Thomas, Director of National Visualization and Analytics Center; Pat Hanrahan, CANON USA Prof. of Computer Science and Electrical Engineering at Stanford Univ.; Nancy Chinchor, Government; George Robertson, Senior Researcher in the Visualization and Interaction Group at Microsoft Research.
Panel 3 - Topic- DIA: Metrics and Measures for Intelligence Analysis Task Difficulty
Chair - Frank L. Greitzer, Battelle, Pacific Northwest Division; Kelcy M. Allwein, DIA
Panelists -John Bodnar, SAIC; Steve Cook, NSA; William C. Elm, ManTech; Susan G. Hutchins, US Naval Postgraduate School; Jean Scholtz, NIST; Geoff Strayer, DIA; Bonnie Wilkinson, AFRL.
Panel 4 - Topic-NGA: Human-Systems Effectiveness: An Integrative Concept for the Intelligence Community
Chair - Keith Masback, Deputy Director of Strategic Transformation, NGA
Panelists - Jeff Baxter, Indep. Consultant; June Skelly, Air Force Rsch Lab; Alan Wade, Intelligence Community and CIA Chief Information Officer
Panel 5 - How Can Predictive Accuracy Be Improved?
Chair -Steven Rieber, Kent Center for Analytic Tradecraft
Panelists - Scott Armstrong, Prof of Marketing, Wharton School, Univ. of Pennsylvania; Mark B. and Jennifer P., methodologists, Directorate of Intelligence, CIA; Sundri Khalsa, author of Forecasting Terrorism; Shelley Kirkpatrick, methodologist, Homeland Security Institute
Tuesday 3 May
8:00 Welcome by Government
8:05 Conference Opening - Dr. Mark Maybury, Co-chair, IA â05
8:10 Keynote Introduction, Mr. Marty Faga, CEO MITRE
8:15 Keynote Address â Dr. Mark Lowenthal, Former Assistant Director of Central Analysis and Production (ADCI A&P)
9:15 Panel One: Joint DIA/NSA (Culture)
10:45 am Coffee Break Exhibits Open
11:00 Paper Parallel Sessions T1A-T1D (3-4 papers)
12:30 Sit down Lunch - Keynote - Maureen Baginski, Executive Assistant Director Directorate of Intelligence, FBI (confirmed) A luncheon ticket will be required for participation in for the Maureen A. Baginski talk.
2:00 Parallel Paper Sessions T2A-T2D
3:30 Coffee Break
4:00 Parallel Paper Sessions T3A-T3D - Panel Two: DHS (Visualization/R&D Needs)
5:30 End Sessions - Exhibits Close
6:30pm Conference Dinner and Best Paper Award
Keynote Speaker: Bran Ferren, Co-Chairman and Chief Creative Officer, Applied Minds Inc
8:00 Conference Opening
8:10 Keynote Introduction â Mark Lowenthal
8:15 Keynote Address -James R. Clapper. Lt Gen, USAF (Ret.), Director, National Geospatial-Intelligence Agency
9:15 Panel Three: DIA (Metrics and Measures)
10:45 am Coffee Break Exhibits Open
11:00 Paper Parallel Sessions W1A-W1D
12:30 Buffet Lunch and Poster & Demonstration Session II (Grand Ballroom B)
2:00 Paper Parallel Sessions W2A-W2D.
3:30 pm Coffee Break
4:00 Paper Parallel Sessions W3A-W3D. Panel Four: NGA. Human-Systems Effectiveness. #31
5:30 End Sessions - Exhibits Close - Conference Closing and Best Poster/Demonstration Award
11:00am Tuesday Paper Session
Thinking Critically - David Moore, NSA - Session Chair (T1A)
- Thinking Critically for Intelligence - David Moore, NSA
Information Extraction I - Dr. Carol Van Ess-Dykema, USG - Session Chair (T1B)
- Tracking Information Extraction from Intelligence Documents - Christopher Welty, J. William Murdock, Paulo Pinheiro da Silva, Deborah McGuinness, David Ferrucci, IBM Watson Rsch Ctr.
Data Fusion and Analysis - Rita Bush, ARDA Session Chair (T1C)
- 21st Century Intelligence Production, Analysis, Access, and Delivery - David N. Drake, Michelle J. Busbee, Daniel B. Gerogosian, Thomas Pedtke Nat'l Air and Space Intelligence Ctr.- Presentation Slides
Visualization - Kelcy Allwein, DIA, Session Chair (T1D)
4:00pm Tuesday Paper Session III
Analytic Collaboration and Training - Jay Hillmer, DIA, - Session Chair (T2A)
- Personal Cognitive Assistants for Military Intelligence Analysis: Mixed-Initiative Learning, Tutoring, and Problem Solving - Gheorghe Tecuci, Mihai Boicu, GeorgeMason Univ., Cindy Ayers, David Cammons, US War College
Information Extraction II - Dr. Salim Roukos, IBM Session Chair (T2B)
- Knowledge Discovery via Content Indexing - Evelyne Tzoukermann and Anthony R. Davis, StreamSage
- Automatic Evidence Extraction - Elizabeth Boschee, Ralph Weischedel, Alex Zamanian, BBN Technologies
- Using Knowledge Extraction and Maintenance Techniques to Enhance Analytical Performance - David Bixler, Dan Moldovan and Abraham Fowler, Language Computer Corp
Panel Five: Steve Rieber "How Can Predictive Accuracy Be Improved? (T2C)
Decision Support and Simulation - Kim Walls, NGA, Session Chair (T3A)
- Combining analogy, intelligent information retrieval, and knowledge integration for analysis: A preliminary report - Kenneth D. Forbus, Larry Birnbaum, Earl Wagner, & James Baker (Northwestern Univ.), Michael Witbrock (Cycorp)
Social Network and Cultural Analysis - Kelcy Allwein, DIA, Session Chair (T3B)
- Supporting Reasoning about Cultural and Organizational Influences in an Intelligence Analysis Decision Aid - Jonathan Pfautz, Adam Fouse, Emilie Roth, Roth Cognitive Engineering Bryan Karabaich, Karabaich Strategic Info. Services Greg Zacharias, Charles River Analytics
Human Information Interaction - Nahum Gershon, MITRE, Session Chair (T3C)
- Building a Human Information Discourse Interface to Uncover Scenario Content - Antonio Sanfilippo, Robert Baddeley, Andrew Cowell, Michelle L. Gregory, Ryan Hohimer, and Stephen Tratz, Battelle-Pacific NW Lat'l Laboratory
11:00am Wednesday Paper Session
Sensemaking and Hypothesis Testing - John Bodnar, SAIC, Session Chair (W1A)
- Sensemaking Support Environment: A Thinking Aid for All-Source Intelligence Analysis Work - Robert G. Eggleston, Air Force Research Lab; Ratna Bearavolu, and Ali Mostashfi SRA International--Presentation Slides
Denial and Deception - Dr. Frank Stech, MITRE, Session chair (W1B)
- A cognitive framework for information gathering with deception detection for intelligence analysis - Eugene Santos, Jr., Qunhua Zhao, Gregory Johnson Jr., Hien Nguyen, Univ. of Conn Paul Thompson,Dartmouth College
- Assessing the Veracity of Criminal and Detainee Statements: A Study of Real-World Data - Douglas P. Twitchell, David P. Biros, Nicole Meek, Judee K. Burgoon, Jay F. Nunamaker Jr., University of Arizona
HLT - Foreign Language - Jim Williams, NMIA, Session Chair (W1C)
- Measuring Translation Quality by Testing English Speakers with a New Defense Language Proficiency Test for Arabic - Neil Granoien, Martha Herzog, Foreign Language Ctr. DLI, Douglas Jones, Wade Shen, Clifford Weinstein, MIT Lincoln Laboratory
- CHINET: a Chinese Name Finder System for Document Triage - K.L. Kwok*, P. Deng*, H.L. Sun*, W. Xu*, N. Dinstl*, P. PengÃÂ§ and J. DoyonÃ¢â¬ *Computer Science Dept., Queens College, City University of New York, Northrop-Grumman Corp. Information Technology, The MITRE Corp.
Video Annotation- Dennis Moellman, DIA, Session Chair (W1D)
Threat Detection - Bob DelZoppo, Syracuse Research, Session Chair (W2A)
- Analysis and Detection of Malicious Insiders - Mark Maybury, Penny Chase, Brant Chiekes, Dick Brackney, Sara Matzner, Brad Wood, Tom Longstaff and Tom Hetherington, Conner Sibley, Jack Marin, Lance Spitzner, John Copeland, Scott Lewandowski and Jed Haile;The MITRE Corp.; National Security Agency; Univ. of Texas; BBN Technologies; Carnegie-Mellon Univ.; Honey Net Consortium; Georgia Institute of Technology; MIT Lincoln Laboratory
Threat Detection (2) - Bob DelZoppo, Syracuse Research, Session Chair Guilt by Association (2) (W2A)
Terrorism and Law Enforcement Steve Dennis, DHS, Session Chair (W2B)
- Establishing Identity Equivalence in Multi-Relational Domains - Jesse Davis, Ines Dutra, David Page, Vitor Santos Costa, Univ. of Wisconsin-Madison
HLT - Speech and Language Processing - Dr. Liz Liddy, Syracuse University, Session Chair (W2C)
- Infrastructure and Systems for Adaptive Speech and Text Analytics - S. Caskey, U. Chaudhari, E. Epstein, R. Florian, J. S. McCarley, M. Omar, G. Ramaswamy, S. Roukos, and T. Ward, IBM TJ Watson Research Ctr.
- Searching Conversational Telephone Speech in Any of the World's Languages - Patrick Schone, Paul McNamee, Greg Morris, Gary Ciany, Stephen Lewis, US Dept. of Defense, JHU/Applied Physics Laboratory, Dragon Development Corp
The Analysis Process and Metrics (2) - Steven D. Rieber, GCU, Session Chair (W3A)
Question Answering - Dr. John D. Prange, ARDA Session Chair (W3C)
- QUery Improvement Elevation Technique (QUIET) - Margaret Knepper, Kevin Fox, Harris Corp. Ophir Frieder, Illinois Institute of Technology
Forecasting - David Moore, NSA, Session Chair (W3D)
POSTERS and DEMONSTRATIONS (excel) or (html)
Mondays Posters and Demonstrations
Athena's Prism: A Strategic Role Playing Simulation for Generating Ideas and Exploring Alternatives - Barry G. Silverman, Richard L. Rees, Jozsef A. Toth, Jason Cornwell, Kevin O'Brien, Michael Johns, Marty Caplan,- Univ. of Pennsylvania
DARPA-sponsored Research in Human Language Technology: The Role of Shared Data -Christopher Cieri, Mark Liberman - Univ. of Pennsylvania;
Applying Semantic Web Technologies for Federation and Semantic Alignment of Disparate Counter-Terrorism Information Sources -Michael Personick, Bradley Bebee, Bryan Thompson, Yunsheng Wang, Mark Jaworowski; -SAIC
On the Scalability to a Terabyte of a Distributed-Index Information Retrieval System versus a Shared-Index Information Retrieval System - Jefferson Heard, David Grossman, Ophir Frieder, Joseph Prokop - Illinois Institute of Technology
Wednesday Posters and Demonstrations
Social Network Analysis
Nexus Topography - Major Steven Marks, Major Matthew Nilson, Captain Tom Meer - U.S. Naval Postgraduate School
Finding Novel Information in Large, Constantly Incrementing Collections of Structured Data - Jaime Carbonell, Cenk Gazen, Chun Jin, Phil Hayes, Aaron Goldstein, Johny Mathew, Dwight Dietrich - Carnegie-Mellon Univ.
Radiant GARNET: Automated Data Mining and Pattern Discovery in Intelligence Data - Brett D. Lapin, Johns Hopkins University
Towards a Model of Pattern Recovery in Relational Data - Benjamin Rode - Cycorp
The "Findability" Quotient - A Key To Making Intel Accessible Maj. Dan Ward, USAF
Learning Concepts from Intelligence Data Embedded in a Supervised Graph -Joseph T. Potts, Diane J. Cook, Lawrence B. Holder and Jeffrey Coble, University of Texas
PROXIMITY: An Environment for Knowledge Discovery in Semantic Graphs - David Jensen- University of Mass
Metrics and Evaluation
Extending Heuer's Analysis of Competing Hypotheses Method to Support Complex Decision Analysis - Marco Valtorta, Michael Huhns, Jiangbo Dang, Hrishikesh Goradia, and Jingshan Huang - Univ. of South Carolina
Analysis of Analysis
Finding Decision Support Requirements for Effective Intelligence Analysis Tools - William Elm, Scott Potter, James Tittle; -ManTech Cognitive SysEng Ctr.;- David Woods, Emily Patterson, Justin Grossman; - Ohio State Univ.
DTB Project: A Behavioral Model for Detecting Insider Threats- Paulo C.G. Costa, Kathryn B. Laskey, Mehul Revankar, Thomas Shackelford, Daniel BarbarÃ¡, Ghazi Alghamdi, Sepideh Mirza;- George Mason Univ.;
Denial and Deception
Counter Denial and Deception and Utility-Theoretic Information Retrieval for Intelligence Analysis - Paul Thompson; Dartmouth College;-Eugene Santos Jr., Qunhua Zhao, Gregery Johnson, Hien Nguyen Univ. of Connecticut
Classification of high explosive type and weight from multi-band imagery - Anthony N. Dills, Ronald F. Tuttle, Glen P. Perram - Air Force Institute of Technology
TUTORIALS and BIOS
(2 page tutorial introduction then presentation slides)
Tutorial 1: Warning Analysis for the Information Age: Rethinking the Intelligence Process -- Presentation Slides
(Dr. John W. Bodnar, SAIC)
QUESTION: How can we rethink our assumptions on intelligence to redefine methodologies to provide strategic warning in an age of WMD and terrorist threat?
HYPOTHESIS: The changing tempo of WMD and terrorist threats has virtually destroyed the ability to provide tactical or strategic warning. New methodologies for warning intelligence need to be developed based on quantum thinking rather than Newtonian thinking.
MODELS: Toward Information Age Intelligence
(1) Modeling the Decision Cycle. Dynamic models for biological systems including organizations can be built using the Decision Cycle developed by COL John Boyd.
(2) Modeling the Target. A methodology for Multidimensional Analysis (MDA) will be presented to model the target that accounts for strategic planning and the need for a two decision cycle planning process to build WMD or CBRNE.
(3) Modeling How We Model. A model for how we think - build and test hypotheses - to provide strategic warning is presented so that we can develop tools to help automate the process and develop training methods for new analysts.
(4) Modeling Ourselves. Recommendations to "reorient the arrows" within the Intelligence Community - both the information flow and leadership interactions - are provided to indicate how the IC can re-think its analytical infrastructure to empower the analyst.
Dr. John Bodnar is a senior biological warfare analyst at SAIC, McLean, VA. His interest in analytical methods and tools for the IC comes from previous experience conducting biological warfare analysis at the Defense Intelligence Agency, analyzing the Revolution in Military Affairs as a Navy Reservist for the Office of Naval Research and the US Naval War College, and researching and teaching bioinformatics at Northeastern University, the US Naval Academy, and Villa Julie College.
Tutorial 2: Tailoring Scenario Analysis for Short-Term Policy Challenges (Bill Fiora and Ken Sawka, Outward Insights)
Until recently, scenarios have been used by both the intelligence community and large corporations as a long-term planning tool to envision plausible future worlds or conditions ten or more years into the future. While this technique is very effective for analyzing the impact of environmental variables on long-term planning, it often fails to resonate with key policymakers who work with shorter-term issues and who need to develop policies they can implement - and see results from - immediately.
In this tutorial, two former CIA analysts who now work with both corporate and government intelligence analysts to improve their performance will discuss how certain aspects of the scenario planning methodology have been adapted for the analysis of more immediate problems and issues. The presentation will examine ways in which scenarios can better anticipate near-term, critical developments without losing the richness and rigor of the scenario approach. We will also examine the input requirements for short-term scenarios, and how this process allows policymakers to better visualize their range of options and possible implications. Finally, we will demonstrate how intelligence analysts can use the process - and the resulting discussion with policymakers - to develop early warning intelligence indicators that will alert policymakers to imminent developments.
Bios: Ken Sawka is a renowned expert in competitive intelligence, scenario planning, early warning systems, and competitive strategy. Ken has been quoted extensively on matters of competitive strategy and intelligence in Time, Investors Business Daily, American Banker, and other prominent journals.
Prior to joining Outward Insights, Ken directed pricing and competitive analysis for the consulting function at Deloitte, one of the world's leading professional services firms. Ken also served as a practice leader in Deloitte's Strategy and Operations Group, managing the delivery of services in strategy development, competitive analysis, and scenario planning for clients in the telecommunications, healthcare, and financial services industries. Prior to joining Deloitte, Mr. Sawka led the competitive intelligence consulting groups at Fuld & Co. and The Futures Group, specializing in intelligence process building and improvement. Mr. Sawka also served eight years as an intelligence analyst with the US Central Intelligence Agency. Ken holds a Masters Degree in International Relations and a Bachelors Degree in Political Science, cum laude, from American University. He is a member of the Society of Competitive Intelligence Professionals - including serving on the Society's Board of Directors from 1999 - 2002 - and the Association of Former Intelligence Officers. Mr. Sawka is also an adjunct faculty member at Simmons College in Boston.
Bill Fiora founded Outward Insights to help companies build and improve analysis capabilities that directly support strategy. He is a recognized expert in competitive intelligence and scenario planning, and writes a Best Practices column for Competitive Intelligence Magazine. He has helped dozens of leading companies build and improve their competitive intelligence capabilities so that they directly support strategy. He has over fourteen years of experience as a practitioner and consultant, having worked at Deloitte Consulting, Zefer, and The Futures Group. He is a former adjunct professor at Tufts University's Fletcher School of Law & Diplomacy.
Prior to launching his consulting career, Mr. Fiora spent six years as a senior analyst at the US Central Intelligence Agency. He holds a Master of Arts degree in International Affairs from Columbia University. He is a member of the Institute of Management Consultants, and the Society of Competitive Intelligence Professionals where he serves as a member of the Executive Board of Directors.
Tutorial 3: Probability and Information: The Essential Facts and Concepts -- Prestation Slides
(Dr. Richard Rohwer, Fair Isaac Corp.)
The mathematical theories of probability and information have ample justification to claim the last word on many pervasive issues arising in the design and application of tools and methods for analysis. They make concrete such abstruse concepts as knowledge, belief, uncertainty, and relevance, and they hold a monopoly on logical consistency. It would be imprudent, at best, to design, evaluate or apply an analysis method or tool without at least lending some thought to it within the conceptual framework of these theories. Yet the mathematically technical nature of the subject makes this all but impossibly difficult for most analysts, managers, and policy makers, and for that matter, more than a few toolmakers. This tutorial aims to enable such people to overcome this barrier well enough to be alert to the salient issues in their own work, and to enjoy pithy interactions with toolmakers and method designers. It will introduce technical terminology and corresponding mathematical notation, but only in so far as is necessary to explain the fundamental concepts of these subjects, how they are interwoven, and how they relate to analysis. Many mathematically important but relatively unenlightening details will be omitted, as will be proofs and derivations.
Richard Rohwer obtained his BS in Physics from Stanford University in 1978, and his PhD in Physics from the University of Texas at Austin in 1985. His thesis concerned quantum effects in early cosmology. After a year of postdoctoral work at the University of Newcastle upon Tyne, in England, he moved to the Center for Speech Technology Research at the University of Edinburgh, Scotland, where he remained through 1991. As a post-doc at Edinburgh, and briefly at Strathclyde University, Glasgow during 1992, his research focused on neural computing and speech recognition, with an emphasis on learning algorithms, sparse distributed memory, and neurodynamics. He then became a Lecturer at Aston University, Birmingham, England, where he remained until 1996. At Aston, his research broadened from neural computing to statistical pattern recognition generally, including Bayesian methods and Information Geometry. In 1996 he moved to Prediction Company in Santa Fe, NM to work on financial time series prediction. He joined the Advanced Technology Solutions Department at HNC Software, San Diego, CA in 1997, where he now holds the title of Director in Analytic Science. HNC became a wholly-owned subsidiary of FairIsaac Corporation in 2002.
At HNC, Dr. Rohwer leads Government-funded advanced research in pattern recognition for diverse applications, and assists with technology transfer. He has been principle investigator on DARPA and ARDA projects in natural language processing, machine vision, and bioinformatics. He is currently PI for the ARDA NIMD SPRITe project. His main research interest is machine cognition, which he views as an information self-organization process.
Tutorial 4: Complexity Management for Decision-Making --- Presentation Slides (Gene Allen and Dr. Jacek Marczyk, Complex Systems Engineering, Inc)
This tutorial will provide an overview of new process to provide a capability for people with complex problems to make better decisions. A means of reducing data that has been successfully used in the past is to have an expert draw a cause and effect diagram referred to as a Fuzzy Cognitive Map (FCM). An FCM was successfully used to help make policy decision on how to address the Apartheid policy of South Africa.
The tutorial will show how FCMs are automatically generated from complex data sets to generate a "Knowledge Map." The foundation for this process has been proven in leading aerospace and automotive design analysis. The tutorial will review how to generate Knowledge Maps from complex data sets and how Knowledge Maps can be used to help make better decisions. Complex problems involve information generated from:
" Physical data sources (sensors, historical data, etc.), or
" Virtual data created from simulations of what-if scenarios.
The process captures all available data and information and incorporates into a Fitness Landscape. The Fitness Landscape is then reduced to a Knowledge Map. Knowledge Maps provide information in a format the can be quickly understood and used by decision makers.
Mr. Gene Allen has established a number of collaborative R&D programs applying computers to design and manufacturing to develop better, safer products. He has worked with Boeing, Lockheed-Martin, Ford, General Motors, Pratt & Whitney, Texas Instruments, Kodak, DARPA, NASA, NIST, DOE labs and others. He recently co-authored "Collaborative R&D: Manufacturing's New Tool" published by Wiley. Previous positions include: Director Collaborative Development at NCMS, Assistant to Senator Robert C. Byrd, and Associate with Booz, Allen & Hamilton. Mr. Allen served as a nuclear-trained Navy officer after graduating from MIT.
Dr. Jacek Marczyk has twenty years of experience in Computer Aided Engineering in the aerospace, offshore, and autmotive industries. In the nineties, he pioneered stochastic simulation techniques and introduced the first commercial software tool into the CAE market for performing realistic large-scale stochastic analysis. The tool has been adopted by prestigeous car manufacturers and aerospace agencies. His significant contribution in the past years has been the propagation of new philosophies and innovative methodologies in the fields of stochastic simulation, uncertainty and risk management in CAE.
Tutorial 5: Bayesian Networks: An Analytic Tool for Predicting Behavior (Dr. Dennis Buede, Mary Lewellyn, Dr. Richard L. Rees, and Dr. Paul Sticha)
APOLLO is a software application in the intelligence community that enables the analyst to reason through a prediction of a Subject's decision making, to identify assumptions and determinant variables, and to quantify each variable's relative contribution to the prediction, producing a graphical representation of the analysis with explicit levels of (un)certainty. The analyst builds Bayesian networks either from scratch or from previous models or templates. Bayesian networks integrate situational information with the Subject's personality and culture to provide a probabilistic prediction of alternative competing hypotheses about actions that a Subject might choose. The personality and culture variables-a unique application of psychosocial expertise-are based on literature and expert consensus.
We provide an overview of the approach to the model development and analysis process, using a computer implementation of the analysis process with a graphical interface.
1. Overview of the modeling method (Bayesian networks)
2. Working through a sample problem.
a. Defining the hypothesized subject behavior
b. Defining and linking to situational variables
c. Adding subject personality information
d. Conducting what-if analyses
e. Conducting sensitivity analyses
3. Using Apollo to develop analyses over time as new data are introduced
4. Exploring applications of the methodology to the work of the participants
Dennis Buede received his Ph.D. and M.S. from the Engineering-Economic Systems Department of Stanford University and his B.S. in Aerospace Engineering from the University of Cincinnati. He is currently a Principal in Innovative Decisions, Inc., which conducts decision and risk analyses. He has been a Professor of Systems Engineering and Engineering Management at Stevens Institute of Technology and Professor at George Mason University.
Mary Lewellyn is a Lead Software Engineer for GH Engineering, Inc., based in Vienna, VA. She is currently pursuing a B.S degree in Computer Science. After completing her degree in Computer Science, she intends to pursue a M.S. in Engineering. She has twelve years of experience in Software Development and System Architecture.
Richard Rees earned his Ph.D. in 1979 in clinical psychology and re-trained in industrial-organizational psychology. He is licensed as an applied psychologist in Virginia. For the past 26 years, he has pursued the application of psychological principles to the world of work, especially the performance of executives and the personality and contextual factors that lead to pro- and anti-social behavior in organizations. Dr. Rees has designed and taught college, doctoral and post-doctoral symposia.
Paul Sticha received a Ph.D. and M.S. in Mathematical Psychology and an M.S. in Mathematics from the University of Michigan. Dr. Sticha manages the Simulation and Modeling Program at the Human Resources Research Organization (HumRRO). His work at HumRRO has focused on the design, development, implementation, and evaluation of mathematical and simulation models to aid in intelligence analysis, military selection and classification, training system design, and business process modeling.
Tutorial 6: Intelligent Information Access --- Presentation Slides
(Dr. Mark Maybury, MITRE)
The purpose of this tutorial is to provide an overview of five intelligent information access technologies: information retrieval, summarization, information extraction, text clustering, and question answering. Participants will learn the current state of the art in these technologies, their applicability for specific analytic tasks, and see live demonstrations of several of the technologies. While the course will not make specific tool recommendations, it will provide participants with a list of internet or commercially available tools in each of these categories.
Bio:As Executive Director of MITRE's Information Technology Division, Dr. Mark Maybury is responsible for the direction of MITRE advanced research and development for intelligence and defense systems. Mark has organised international conferences, given tutorials, and published over fifty articles in the area of language generation, multimedia presentation, text summarization, intelligent information retrieval and analysis. Mark 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) and Advances in Knowledge Management: Classic and Contemporary Works (MIT Press, 2001) and co-author of Information Storage and Retrieval: Theory and Implementation. 2nd Edition (Kluwer Academic, 2000) and co-editor of Knowledge Management (MIT Press 2000). Mark serves on the Board of the Object Management Group, Treasurer of ACM SIGART, and member of the Steering Committee for ACM IUI. Mark received his M.Phil. in Computer Speech and Language Processing (1987), an MBA from RPI (1989), and his Ph.D. in Artificial Intelligence (1991) at Cambridge University, UK. Prior to joining MITRE, Dr. Maybury served as a US Air Force Officer at Rome Air Development Center.
Afternoon Tutorial Session 1 pm - 5 pm
(Lunch 12-1 and Coffee Break at 3:30 am)
Tutorial 7: Integrating Methods and Tools to Counter Denial and Deception (Ed Waltz, BAE Systems)
Analysts face severe analytic challenges when seeking to counter a wide range of targets (e.g. enigmas, concealed facilities, covert programs, secured leadership, deep-hide weapons) that employ denial and deception (D&D) to avoid discovery, evade surveillance and deny description. This tutorial introduces the principles of D&D employed to counter intelligence collection and analysis, before describing the methodologies and associated analytic tools that may be employed in an effort to counter a variety of these activities. Participants will gain an understanding of the principles of D&D, the challenges to collection and analysis, and the prominent analytic methods and tools that may be integrated to focus on the most difficult targets to counter D&D. The tutorial is structured according the following outline: Survey of the Classic D&D process models with examples to illustrate the models, implementing the processes, applying inferential and exploratory tools; Underlying Principles of Denial and Deception; Elements of Counter D&D (CDD): Collection, Processing and Analysis Strategies; CDD Processing and Analysis Methods; Integrating the Methods and Tools.
Ed Waltz is the Chief Scientist, Intelligence Innovation Division of BAE Systems Advanced Information Technology, where he leads hard intelligence target research. He has led several counter D&D studies and tool developments over the past decade for different agencies of the IC. He holds a BSEE from the Case Institute of Technology and an MS in Computer, Information and Control Engineering form the University of Michigan, and over 35 years of experience in developing and deploying signal processing, data fusion and analytic technologies for C2 and intelligence. He is the
author of Knowledge Management in the Intelligence Enterprise (Artech 2003), Information and Warfare Principles and Operations (Artech 1998), coauthor of Multisensor Data Fusion (Artech 1990), and coeditor of Multisensor Data Fusion (Kluwer 2001). He is a recipient of the DoD Joseph Mignona Data Fusion Award (2004), and became a Veridian Technology Fellow in 2002. Mr. Waltz is coauthor of the technical text Counterdeception Principles and Applications for National Security, forthcoming from Artech House in the fall of 2005.
Tutorial 8 Threat and Risk Analysis (TRA): The Intelligence Analyst's (IA) Contribution (Henry Doucette
In this post 9/11 environment, the focus on security and force protection issues has significantly increased the demand for suitably developed and accurate intelligence. A residual product that has often been touted as essential is the Threat and Risk Analysis (TRA) frequently produced by security/force protection advisors with intelligence branch input. Although the Intelligence Community continues to develop innovative and effective methods of analyzing and preparing a well developed product, the end users (ie Force Protection advisors) have often, subjectively relegated the intelligence analysis ineffective or at least reduced its significance in favour of a more mechanical matrix driven format. Although there are several TRA methods, forms and systems currently in use, they generally appear to be lacking in the proper employment of the intelligence cycle and the assimilation of analyzed threat intelligence. The resulting risk analysis therefore becomes questionable and the consequences hazardous. The aim of this tutorial is to review the purpose of the TRA, clarify existing government policy and terminology, illustrate and discuss the role and responsibilities of the intelligence analyst (IA), and demonstrate an "Intelligence Based" process and format that will provide the Commander/manager with a straightforward and accurate report.
Henry Doucette served over 27 years with the Canadian Regular Military, 14 of which with the Intelligence Branch and 10 with the Military Police and Security Branch and continues to serve as an Intelligence Analyst for the Department of Defence in Ottawa. He has been studying the use of TRAs and how they are applied for over 15 years, originally out of necessity when attempts at finding existing policy failed. In addition to studying criminology and political science, at a Bachelor level, he is currently working on a tailored Masters Program (Canadian Military College) focusing on war studies, specifically how intelligence is, and has been used throughout history, innovative regional intelligence analytical techniques and most recently the asymmetric threat analysis techniques. He has prepared TRAs for operations such as; Kabul, Kandahar, Istanbul, DRC, amongst other operations, as well as static sites in Kingston, Trenton and the military wing of the Canadian embassy here in Washington. He has worked as an instructor at both the Canadian schools of Military Intelligence and Military Police/ Security. He has a long list of related training conducted in Canada, the US and the UK.
Tutorial 9: Role Playing Strategy Games for Intelligence Analysis (Prof. Barry G. Silverman, University of Pennsylvania, Dr. Richard L. Rees, Jozsef A. Toth, Institute for Defense Analyses) - [Maximum of 30 attendees]
With intelligence analysis, one is continually trying to get people to: 1) see the alternative hypotheses of what might be motivating various subjects of interest; 2) break out of thought silos and biases; and 3) explore the potential effects of diverse actions straddling economic, psychological, cultural, political, diplomatic, military, and other domains. Games often can be the best way to provide such thinking shifts. As a result, this tutorial will introduce game design in terms of a number of role playing game (RPG) ideas and videogame examples that illustrate intelligence-analytic thinking. The participants will then have a hands-on session playing a world strategy game that is a publicly releasable derivative of one developed for the U.S. Government. They will first play against each other in a paper-based version called PolicyMaker (soon to appear at stores). Next, participants will interact with a computer-based variation of PolicyMaker-called LeaderSim-including a set of agent-based software opponents. Lastly, in a learn-by-doing session, we will break into teams and brainstorm new role playing games that may be of value to the participants.
Dr. Barry G. Silverman is Professor in Engineering (ESE & CIS), in Medicine, and in Wharton (OPIM) at the University of Pennsylvania where he is also Director of the Ackoff Center for Advancement of the System Approach (ACASA). He holds the BSE ('75), MSE ('77) and PhD (also '77) all from the University of Pennsylvania, is a Fellow of IEEE, AAAS, and the Washington Acad. of Science, and sits on the board of several organizations and journals in the intelligent systems fields. The focus of his research has largely been on aesthetic and cognitive engineering of embedded game-theoretic agents that can help humans improve their learning, performance, and systems thinking in task-environments. Over the years, his lab has produced or is in the process of creating patient and human physiology simulations; a terrorist campaign and crowd simulator; numerous autonomous and emergent agent tools; several distributed, computer-mediated, human-to-human collaborative systems; 3 role playing games (RPGs); the AESOP interactive fiction game generator; and the PMFserv human modeling system for enhancing the realism of game AI. As a result of all this, Barry is the author of over 120 articles, 13 books/proceedings, over 100 tech reports, 9 copyrighted software systems, a boardgame, and several research and teaching excellence awards.
Dr. Richard Rees earned his Ph.D. in 1979 in clinical psychology and re-trained in industrial-organizational psychology. He is licensed as an applied psychologist in Virginia. For the past 26 years, he has pursued the application of psychological principles to the world of work, especially the performance of executives and the personality and contextual factors that lead to pro- and anti-social behavior in organizations. Dr. Rees has designed and taught college, doctoral and post-doctoral symposia.
Dr. Jozsef A. Toth is a Research Staff Member at the Institute for Defense Analyses, Alexandria VA, and has been involved in the research and development of cognitive modeling, simulation, and theory for the past twenty-five years. Since 1998, Dr. Toth has been engaged in the exploitation of cognitive, AI, and organizational theory in the development and evaluation of tools and business models to support training, analysis, planning, and operations in the next generation of counterterrorism, counterintelligence, and counterproliferation. He is presently working with the Joint Intelligence Task Force-Combating Terrorism at DIA and the Missile Defense Agency in the domain of crisis action planning regarding the human and social factors of collaborative software.
Tutorial 10 Observations on Cognitive Factors in Design
(Prof. Thomas T. Hewett, Drexel University)
"You were issued a memory at birth, but nobody gave you an owner's manual." Dr Tom Hewett is a Cognitive Psychologist/Cognitive Engineer active in the field of Human Computer Interaction. The focus of this tutorial is cognitive issues in memory and problem solving and their implications for the design of effective working environments for others or for one's self. Unlike brief overviews of Gestalt principles or short-term memory that frequently appear in design material, Tom provides a review of some basic research and numerous "live" experiments to drive the issues home. These practical exercises are useful to anyone involved with doing, managing, teaching or promoting interaction design. They will also appeal to those wanting to gain greater understanding of the underlying psychology or wanting to take a hand in design of their own working environment. Individuals looking for a review of the latest research on memory or problem solving should look elsewhere. Web or UI designers looking for "instant" guidance that could be applied to next week's project will also be disappointed. The approach is more reflective and needs to be absorbed over time. In fact, it is expected that participants will make the effort to occasionally re-visit the material.
Bio: Tom Hewett is Professor of Psychology and Professor of Computer Science at Drexel University, Philadelphia, PA, USA. His teaching includes the Psychology of Human Computer Interaction Design and the Psychology of Problem Solving and Creativity. Tom is a published software author and has written on the use of computers in teaching. His current research focuses on understanding the computing support needs of expert knowledge workers.
Tutorial 11: Event and Temporal Awareness for Intelligence Analysis --- Presentation Slides
(Prof. James Pustejovsky, Brandeis University, Prof. Inderjeet Mani, Georgetown University)
We propose a half-day tutorial for analysts on the topic of temporal and event aware intelligence analysis tools.
Natural language data that analysts encounter is rife with references to past and planned events. Without a robust ability to identify and temporally situate events of interest in natural language data, the real importance of the information can be missed. A new generation of language analysis tools have emerged that can allow an analyst to rapidly browse through a knowlege base of events detected in natural language data. These tools are able to temporally organize these events in terms of their time of occurrence. Such tools can be integrated with visualization, summarization, question answering, and link analysis systems to help the analyst work more efficiently in large event-rich information spaces.
The objectives of this tutorial are: 1. To learn about the importance of text-derived temporal and event-based information for analysis 2. To become familiar with programs that enhance event and temporal awareness 3. To learn the strengths and limitations of these technologies. 4. To identify the opportunities of exploitation of such technologies and the information derived using them.
The tutorial will offer a framework in terms of which the space of such tools can be analyzed. We will discuss annotation schemes(e.g., TimeML, a markup language for events, times, and orderings funded by ARDA), as well as standards to support integration and interleaving of temporally aware browsing, searching, querying, and visualization capabilities. We will survey both research and COTS tools relevant for this task.
Bios:James Pustejovsky is Professor of Computer Science and Director of the Laboratory for Linguistics and Computation at Brandeis University. Dr. Pustejovsky conducts research in the areas of computational linguistics, lexical semantics, knowledge representation, and information retrieval and extraction. In 2002 and 2003, he was PI and Co-PI of two ARDA-funded NRRChosted workshops associated with the AQUAINT Research effort, both of them focusing on temporal and event awareness for language texts. Currently, he is PI of an ARDA grant on the same topic. He has participated in numerous DARPA and NSF efforts in Knowledge Extraction and Natural Language Engineering, including the MUC and TIPSTER projects. Pustejovsky was the founding chair of the ACL's Special Interest Group on the Lexicon, SIGLEX, and was editor of "Squibs and Discussion" in the journal Computational Linguistics from 1993-1996. He has organized and chaired numerous international workshops and conferences on language technologies, computational semantics, and linguistics, and was chair of the 1994 ACL general conference in Las Cruces, New Mexico. He is author or editor of numerous books on semantics and natural language processing, including: The Language of Time: A Reader (2005), Oxford University Press, Language and the Multiplicity of Meaning (2005), MIT Press, The Lexicon Reader (forthcoming), MIT Press, Lexical Semantics and the Problem of Polysemy (1997), Oxford, Corpus Processing for Lexical Acquisition (1996), MIT Press, The Generative Lexicon (1995), MIT Press, Semantics and the Lexicon (1993), Kluwer, Lexical Semantics and Knowledge Representation (1992), Springer.
Inderjeet Mani is Associate Professor of Linguistics and Head of the Computational Linguistics Program at Georgetown University in Washington, DC. Prior to joining Georgetown full-time in 2003, he was Senior Principal Scientist at the MITRE Corporation, which he joined in 1992. He has published three books (including two on Automatic Summarization) and over fifty papers; in the temporal area, in addition to publishing papers, he has co-edited the book The Language of Time: A Reader (Oxford University Press, to appear), and co-edited a special issue of the ACM Transactions on Asian Language Processing on Temporal Information Processing. He was Principal Investigator on the DARPA TIDES program, leading the project which developed the TIMEX2 temporal annotation scheme. Subsequently, he participated in the ARDA-funded Summer'2002 workshop on Time and Event Recognition for Question Answering Systems,which developed TimeML, and was co-PI on the ARDA-funded workshop on TANGO: TimeML Annotation Graphical Organizer. He co-organized the workshop on Spatial and Temporal Information Processing (ACL'2001) and on Annotation Standards for Temporal Information in Natural Language (LREC'2001). He has also taught the use of annotation tools in courses at Georgetown University. Dr. Mani has served on the Editorial Board of Computational Linguistics (2002-2004) and on the Program Committees of conferences such as ACL, HLT, AAAI, IJCAI, and the International Conference on Text Generation. He was tutorial chair for ACL'2004, and has chaired numerous workshops.
Tutorial 12: Advanced data mining, link discovery and visual correlation for data and image analysis --- Presentation Slides (Prof. Boris Kovalerchuk, Central Washington University)
Data mining and link discovery have obvious value and great potential for intelligent analysis (IA). It is important that abilities as well as limitations of these methods be well understood in the Intelligence Community. Traditional numeric statistical data mining methods have relatively limited applicability in IA, because data are often not numeric and have a very asymmetric pattern representation. For instance, there are only a few terrorism messages in the stream of normal ones. New relational data mining and link discovery methods have much greater potential to address these challenges. This tutorial describes important advantages of relational methods over traditional methods for discovering, linking and integrating a variety of data including financial and geospatial data. Integration of imagery is another domain where relational methods have important advantages for dealing with images of different resolutions, sensor modalities, viewing angles, and geometric projections. Participants will learn the state of the art in these technologies.
Bio: Dr. Boris Kovalerchuk is a professor of Computer Science at Central Washington University and Director of the Imaging Laboratory at the University (http://www.cwu.edu /~Imaglab). He holds a Ph.D. in Computer Science and Applied Math, and an M.S. in Mathematics. Dr, Kovalerchuk has 30 years of research and teaching experience, and published two books on Data Mining, and Visual and Spatial Analysis (Kluwer, 2000, Springer, 2005), 5 book chapters and more than 80 papers on the subjects of closely related to this tutorial. He is deeply involved in the fields of data mining, link discovery, image integration, and analysis. Dr. Kovalerchuk has organized special sessions, delivered tutorials, presented new research at major conferences, such as SPIE Conferences on Defense and Security. He is a principal investigator on three research projects funded by the Intelligence Community (ARDA/NGA, http://www.ic-arda.org/InfoExploit/aquaint/gi2vis /projects/cffvcaia.html, http://www.cwu.edu/~Imaglab). In this research he works with partners from national laboratories and industry.