MapReduce based frequent itemset mining algorithm on stream data

Offers on e-commerce websites have been mostly a decision made by companies for advertising or clearing stocks. KAAL algorithm was used on sample transaction data to generate frequent itemsets. These frequent itemsets will give an idea of offers to be made on purchase of base items. With advent of internet, the amount of data being generated by business processes is growing exponentially. This paper makes use of Hadoop MapReduce framework to generate association rules on transaction data stream. Offers are suggested spontaneously as the frequent itemsets are being generated at runtime. The paper concludes that the execution time has a linear relationship with number of transactions per batch. It was found that increase in stock size did not have much impact on execution time. Execution time is also inversely proportional to number of nodes.