Mining constraint based association rules on Pest data using closed frequent items

Abstract

One of the very important tasks in data mining is discovering association rules. Agricultural data are voluminous and thus generating all frequent item sets requires large memory as well as takes high computational time. A new approach for the analysis of soybean pest data has been proposed using an algorithm of frequent closed item sets to reduce computational time, and to overcome memory explosion problems. Frequent closed patterns and redundant rules are identified using frequent closed item set mining with intervals. The algorithm produces highly important or redundant rules compared with traditional algorithms. Proposed algorithm requires less memory and as well takes less time compared to Apriori algorithm that also generates frequent closed item sets in such dense dataset.