Analogy, Knowledge Integration, and Task Modeling Tools for Intelligence Analysts
Sponsor: National Science Foundation, Knowledge Discovery and Dissemination Program
Principle Investigators: Kenneth D. Forbus, Lawrence Birnbaum, Douglas Lenat (Cycorp)
Project Summary: Intelligence analysts are the knowledge workers most directly involved in our nation's defense. Analysts must sift through massive amounts of data, using perspective gained from history and experience to pull together from disparate sources the best coherent picture of what is happening. Information Technology research has the potential to create new software tools that could aid analysts in several ways:
- Analysts make heavy use of precedents and analogies. This sometimes leads to vital "trans-logical" leaps. Because of
fundamental human cognitive limitations, it also sometimes leads to false
analogies, where the matches are too superficial, and to missed opportunities,
where the matches are too obscure for unaided human reasoning to uncover.Precedents and analogies therefore both
increase and reduce the quality of the analytic product; a suite of
knowledge-based software power tools could help the analysts recognize the
deeper or less obvious analogies, could help them apply these analogies to the
current situation, and could help them reject superficially plausible but
useless analogies and precedents more rapidly.
- Analysts make heavy use of scenario generation, both to interpret current reported data and to
project plausible future events to expect or guard against. Because of fundamental human cognitive
limitations, the first one or two plausible interpretations (or predictions)
are likely to "lock" us into states where it is difficult to generate all the
other likely interpretations. Knowledge-based software tools can mechanically generate (and evaluate
for consistency and plausibility) additional interpretations and scenarios, and
present them to analysts to help break this logjam.
- Analysts are responsible for being aware of and implicitly knowing the entire contents of immense corpora of data, whose volume far exceeds individual human cognitive capabilities. What the media today calls "connecting the dots" — and blames intelligence analysts for failing adequately to do — is essentially a process of deducing logical consequences of several needles spread across multiple haystacks. Knowledge-based software tools can augment information-retrieval tools for finding relevant pieces of information, and moreover for semantically joining or otherwise interrelating them in the "connect the dots" fashion, to produce conclusions of interest.
Current technology is capable of providing some of this functionality, but in a limited and piecemeal manner. Knowledge-based systems offer fine-grained and logically coherent inferences and hypotheses — deduction and induction — but only when a sufficiently large fraction of all relevant information is both present and represented precisely (e.g., in formal logic, if-then rules, etc.). Analogical reasoning systems offer the prospect of "thinking outside the box" — but again depend upon structured representations. IR (Information Retrieval) systems can handle the quantity and diversity of unstructured information that exists in the world, but cannot generate new inferences or hypotheses.
We are integrating and extending these three technologies to create power tools for intelligence analysts. We expect this to have two simultaneous effects: to advance the state of research in each of our areas, and to lay the groundwork (including prototypes) for developing the broad yet smart software assistants intelligence analysts need.