Obtaining fuzzy membership functions and rule sets in fuzzy expert systems with AHP

Discusses the use of the analytic hierarchy process (AHP) for obtaining membership functions and rules in constructing fuzzy expert systems. The issues are discussed in the context of fuzzy expert systems for cost estimation. AHP is compared with a less rigorous interview approach by means of an experiment using data obtained from expert construction cost estimators. Some of the likely advantages and disadvantages of an AHP approach are discussed.