Milling typical association rules with linguistic terms

Using the given minimum support and minimum confidence, a large set of association rules with linguistic terms can be discovered and therefore it is difficult for the users to obtain the most interesting ones. An algorithm for mining typical association rules with linguistic terms is presented. In this algorithm, the mined association rules with linguistic terms are grouped by the same consequent, and the association rules with linguistic terms in each sub-group are clustered by the dissimilarity between the rules. The association rules with respect to the medoids are selected as the typical association rules with linguistic terms. These typical association rules with linguistic terms are the most representative rules in the set of association rules with linguistic terms.