A Novel Way for Mining Frequent and Interesting Patterns using Genetic Algorithm

Over the years, the process of finding interesting association rules has become a keystone in adept decision making. Among various data mining techniques, Association Rule Mining is mostly used for finding interesting associations among different products in a transactional database. However, mining association rules wind up in finding plethora of rules and hence finding the most ?interesting? and ?optimal? rule becomes a irksome task using generally used Apriori algorithm. However Apriori Algorithm uses Conjunctive nature of association rules, and single minimum support threshold to reveal the interesting rules. But only these factors don't seem sufficient to unearth the interesting association rules efficaciously. Hence, in this paper we have introduced a entire distinct approach for finding much optimized association rules using numerous and varied quality factors like support, confidence, comprehensibility and interestingness. The demonstration performed on copious datasets shows the much improved performance than Apriori algorithm.