Selecting Perfect Interestingness Measures by Coefficient of Variation Based Ranking Algorithm

Abstract

Ranking interestingness measure is an active and essential research domain in the process of knowledge discovery from the extracted rules. Since various measures proposed by many researchers in various situations increases the list of measures and these are not able to use as a common measures to evaluate the rules, knowledge finders are not able to identify a perfect measure to ensure the actual knowledge on database. In this study, we presented about a ranking method to identify a perfect measure, which also reduces the number of measures. Ranking will be done by increasing order of Coefficient of Variation (CV) and not applicable measures are eliminated. Also we introduced heuristic association measures, U cost, S cost, R cost, T combined cost and ranked with existing measures using CV based ranking algorithm, our measures are placed in better position on ranking, compared with the existing measures.

Classification of objective interestingness measures

“…3,4,6,7,9,11,13,15,16,20,22,24,26,27,29,34,35,36,42,45,46,58,63,79,80,95 IND. 4,7,11,14,25,26,33,34,35,42,45,46,47,48,49,50,51,53,54,62,64,65,69,71,74,75,76,77,78,79,81,…”