Knowledge-based fusion of Formets: discussion of an example

This research project for decision support and situation awareness goes beyond classical systems for command and control. As enhancement to tactical situation display, a tool based on situation independent fusion methods is developed. Its rules and templates are given by background knowledge and a priori data. We analyze formatted qualified command language (e.g. Formets) for its suitability to represent information reported by natural language messages. Since we assume our problem mostly to be of a matching nature, this kind of pattern recognition results in the design of templates for aggregation of corresponding information. A selection of intelligence reports from an ambush situation serves as an example. The importance of context information as integral part of the knowledge-based architecture for information fusion is shown.