Prediction of customer buying nature from frequent itemsets generation using Quine-McCluskey method

Finding frequent itemsets is one of the most important fields of data mining and also finding the rules associated with the itemset is another important ground in association rule mining. The process of finding new techniques to reduce candidate generation in order to generate frequent itemsets efficiently is a challenging task and generating rules in order to know the purchase behaviour of the customer to improve the business. In this paper, an efficient and optimized algorithm called Customer Purchase Behaviour (CPB) has been introduced for finding frequent itemsets using minimum scans, time and memory and the rules are generated. The subset invention process is used for the generation of frequent itemsets which reduces the intermediate tables. This approach reduces main memory requirement since it considers only a small cluster at a time. The purchase behaviour of the customer can be easily judged by using the Quine-McCluskey method. This algorithm is very efficient because of redundancy elimination and rule generation then compared with Apriori, Cluster-Based Bit Vector Mining for Association Rule Generation (CBVAR) and Improved Cluster-based Bit vector mining algorithm for Frequent Itemsets Generation (ICBV).