Building decision aids: Exploiting the synergy between decision analysis and artificial intelligence

Abstract
A practical approach to decision aid development is outlined. This approach systematically exploits both the problem-structuring techniques of decision analysis and the incrementally modifiable software architectures found in artificial intelligence. The approach advocates the use of decision analysis modeling techniques for the initial structuring of expert knowledge. At the same time it advocates the use of artificial intelligence software architectures that separate a domain-specific knowledge base from general problem-solving procedures. Two specific aid architectures that reflect specific instantiations of this approach are shown. These aid architectures are drawn from two ongoing decision-aid development efforts.