Probability Concepts for an Expert System Used for Data Fusion

Probability concepts for rule-baaed expert systems are developed that are compatible with probability used in data fusion of imprecise information Procedures for treating probabilistic evidence are presented, which include the effects of statistical dependence. Confidence limits are defined as being proportional to root-mean-square errors in estimates, and a method is outlined that allows the confidence limits in the probability estimate of the hypothesis to be expressed in terms of the confidence limits in the estimate of the evidence. Procedures are outlined for weighting and combining multiple reports that pertain to the same item of evidence. The illustrative examples apply to tactical data fusion, but the same probability procedures can be applied to other expert systems KNOWLEDGE-BASED EXPERT SYSTEMS are a class of computer programs intended to serve as consultants for decision making. These programs use a collection of facts, rules of thumb, and other knowledge about a limited field to help make inferences in the field. They differ substantially from conventional computer programs in that their goals may have no algorithmic solution, and they must make inferences based on incomplete or uncertain information. They are called expert systems because they address problems normally thought to require human specialists for solution, and This research was supported by the Lockheed Independent Research Program An early version of this paper was presented at the Seventeenth Annual Asilomar Conference on Circuits, Systems and Computers, Pacific Grove, California, October 31-November 2, 1983. knowledge-based because researchers have found that amassing a large amount of knowledge, rather than sophisticated reasoning techniques, is responsible for the success of the approach. Advantages of an expert system are that is can be designed to supply one or more hypotheses to the user, request additional information from the user, explain to the user the reasons for the hypotheses or for the requests for additional information, and allow the addition or deletion of knowledge and rules without extensive reprogramming. A recent survey article by Gevarter (1983) discusses expert system applications in areas such as medical diagnosis, geology, and computer configuration analysis and evaluates the limitations of current systems. Limitations exist on the current use of expert systems because formalizing the knowledge of experts is a difficult task; building the system is laborious and timeconsuming; operation is effective only in a relatively limited field; and degradation is not always graceful when problems are outside of the limited field. Newer expert systems are being developed that find ways around these limitations, and future use and growth of expert systems should increase. Duda and Shortliffe (1983) discuss current research in expert systems, while the book edited by Barr and Feigenbaum (1982) presents more detailed material. This article is concerned with the data fusion aspect of expert systems. The correlation and fusion of information from sensor systems is becoming increasingly important for THE AI MAGAZINE Fall 1984 55 AI Magazine Volume 5 Number 3 (1984) (© AAAI)