Fuzzy retrieval in case-based decision support systems

This paper describes a technique which determines suitable product offers from data bases using the theory of fuzzy sets. Each of the decision criteria is assessed by the customer. His preferences are transformed into membership functions. One gets a suitable offer by maximizing a function which results from the composition of the individual membership functions. We propose the Werners operator for this composition because: (i) it permits the search for a suitable offer to be performed by solving a linear Boolean optimization task and (ii) it reacts more sensitively to customer wishes than the minimum operator. This approach is also feasible for solely verbal criteria and data bases of bigger size.