Expert systems as automated decision aids

A computer expert system is an automated tool for representing and manipulating the knowledge and problem solving techniques of human experts. The expert system differs from conventional computer programs in that its tasks may have no algorithmic solutions and the system must make decisions based on incomplete or uncertain information. The expert system uses a collection of facts, rules of thumb, and other knowledge about a particular topic area in order to generate hypotheses and make inferences. In the field of signal and information processing, expert systems are a way to bring together and enhance other, more traditional, approaches such as filtering, decision theory, and engineering judgement. In this paper a tutorial exposition is presented of an example expert system for correlation and classification of messages from electronic intelligence.