Abstract : BAE Advanced Information Technologies (AIT) has developed an Information Interpretation and Analysis Toolkit (I2AT) that provides a set of analytic capabilities that fit within and augment a knowledge management system. The I2AT tool will address the challenge of effectively analyzing vast amounts of information with limited analyst manpower by focusing analyst attention on available pieces of information that produce significant changes in the assessment of the situation and identifying additional information that has the potential to do so. The central capabilities of the I2AT are as follows: interpretation, false information detection, and data needs generation. The I2AT software uses Bayesian Networks, a probabilistic modeling framework, augmented with a suite of algorithms for analyzing hypotheses, data, and value of additional information. Automated generation of potential interpretations will help reduce the time needed to assimilate and act on new information. At the same time, flagging new data that is inconsistent with existing information will identify potential errors in the knowledge acquisition process or adversary attempts at deception, enabling preemptive correction before erroneous interpretations precipitate actions. I2AT implements an error model for information sources into the Bayesian Network that allows it to reason about the correctness of particular reports and what new evidence would best resolve ambiguities. Within a knowledge management framework, data that are tagged as potentially deceptive can be tracked to determine what analytic results need to be called into question. Finally, I2AT will be able to determine which additional information would have the greatest potential to explain observed inconsistencies.
[1]
Finn V. Jensen,et al.
Bayesian Networks and Decision Graphs
,
2001,
Statistics for Engineering and Information Science.
[2]
Qin Liao,et al.
Online Learning of Bayesian Network Parameters
,
2008,
2008 Fourth International Conference on Natural Computation.
[3]
Judea Pearl,et al.
Probabilistic reasoning in intelligent systems - networks of plausible inference
,
1991,
Morgan Kaufmann series in representation and reasoning.
[4]
David J. Spiegelhalter,et al.
Probabilistic Networks and Expert Systems
,
1999,
Information Science and Statistics.
[5]
Simon Parsons,et al.
Addendum to "Current Approaches to Handling Imperfect Information in Data and Knowledge Bases"
,
1996,
IEEE Trans. Knowl. Data Eng..
[6]
D. Schum.
The Evidential Foundations of Probabilistic Reasoning
,
1994
.