Linguistic summarization of data sets using association rules

We discuss linguistic summaries of databases introduced by Yager. Starting with our previous work we propose some extensions to the form of a linguistic summary. The new form still fits the scheme of an association rule. However, in comparison to the previous approaches, the use of a range of linguistic values as items and their linguistically quantified aggregation is allowed. An algorithm for mining such association rules within this framework, via the authors´ FQUERY for Access package, is presented. Moreover, we show that Zadeh´s idea of protoforms, and their hierarchies, can be employed to represent various types of linguistic summaries.