A Parallel Approach to Combined Association Rule Mining

Data Mining carried out using traditional methodologies of Support-Confidence framework and Association Rule Mining yield an enormous number of inefficient rules or patterns in a certain amount of time. In this paper, a parallel approach to Combined Mining has been implemented that not only generates rules which are "Actionable" but also does so in a time period that is lesser than that of the traditional approach. These implementations are carried out on datasets at different locations consisting of multiple related data items and are independent of each other. The results of an Apriori algorithm is fed as an input to Combined Mining so as to generate more useful patterns for the process of business decision making.