A Dialectic Approach to Intelligence Data Fusion For Threat Identification
Small Business Information
Bevilacqua Research Corp.
2227 Drake Ave, Suite 204, Huntsville, AL, 35802
Andrew T. Bevilacqua
AbstractThe overall objective of this program will be to demonstrate the feasibility of a Dialectic Neural Network (DNN) architecture for use as an intelligent information processing engine for threat identification. To achieve this goal, several technical objectives have been identified that define, integrate and demonstrate the DNN architecture within the Army's Phoenix C41 software which will be used as a demonstration testbed for Phase I. The DNN architecture uses a matrix of neural networks and conceptual graphs. The graphs store knowledge/information in an object-oriented format. When retrieved, through a question to the decision aid, the stored knowledge is passed to the neural networks which are trained to associate the knowledge elements and form an answer in the desired format. A successful demonstration in Phase I will allow the DNN to be further developed for use as a threat assessment tool that can be applied to a wide variety of problems where knowledge handling, data fusion or human decision aiding are needed. Preliminary discussions with commercial product manufacturers indicate that the commercial applications of the DNN-based architecture are numerous and that the ability of the DNN to store and use cognitive (expert) knowledge, makes this particular architecture much more robust than rule-based or statistically-based methods. In addition to military information processing for the battlefield, the TIPS architecture can be used in the emergency management industry, for route planning, intelligent highway vehicle systems (IHVS) applications at the Federal Highway Administration, and on the internet as a tool to provide intelligent routing and handling of the growing volumes of information available in that forum.
* information listed above is at the time of submission.