Linguistic Database Summaries Using Fuzzy Logic: towards a Human-consistent Data Mining Tool

We discuss an approach to fuzzy linguistic summaries of data (bases) in the sense of Yager, i.e., for instance, if we have a (large) database on employees, and we are interested in a relation between the age and qualifications, then it may be summarized by, say, “most young employees are well qualified”. We present the derivation of such linguistic summaries in the context of Zadeh’s computing with words and perceptions paradigm, and consider his recent idea of a protoform to define and handle more general forms of summaries as proposed by Kacprzyk and Zadro ny. We present an implementation for a small to medium computer retailer, and show how data from the internet can qualitatively enhance the summarization results.