Situational awareness is critical on the modern battlefield. A large amount of intelligence information is collected to improve decision-making processes, but in many cases this huge information load is even decelerating analysis and decision-making because of the lack of reasonable tools and methods to process information. To improve the decision making process and situational awareness, lots of research is done to analyze and visualize intelligence information data automatically. Different data fusion and mining techniques are applied to produce an understandable situational picture. Automated processes are based on a data model which is used in information exchange between war operators. The data model requires formal message structures which makes information processing simpler in many cases. In this paper, generated formal intelligence message data is visualized and analyzed by using the self-organizing map (SOM). The SOM is a widely used neural network model, and it has shown its effectiveness in representing multi-dimensional data in two or three dimensional space. The results show that multidimensional intelligence data can be visualized and classified with this technique. The SOM can be used for monitoring intelligence message data (e.g. in purpose of error hunting), message classification and hunting correlations. Thus with the SOM it is possible to speed up the intelligence process and make better and faster decisions.

Date Published: 19 May 2009
PDF: 10 pages
Proc. SPIE 7352, Intelligent Sensing, Situation Management, Impact Assessment, and Cyber-Sensing, 735204 (19 May 2009);

Published in SPIE Proceedings Vol. 7352:
Intelligent Sensing, Situation Management, Impact Assessment, and Cyber-Sensing
Stephen Mott; John F. Buford; Gabriel Jakobson, Editor(s)