Abstract

Abstract:Data mining is that the extraction of hidden prognostic data from massivedatabases, that could be a powerful new technologywith nice potential to assist firms likewise as analysis specialise in the foremostnecessary data in their information warehouses. Dataprocessing tools predict future trends and behaviours, permitting businesses to create proactive, knowledge-driven choices. Frequent Itemset Mining is one among the classical data processing issues in most of the information mining applications.It needs terribly massivecomputations and I/O traffic capability.Resources like single processor’s memory and processor area unitterribly restricted, thatdegrades the performance of algorithmicprogram. During thispaper we've planned associate degree algorithmicprogram which can run onHadoop – one among the recent hottest distributed frameworks that chiefly specialise inMapreduce paradigm. The plannedapproach takes into considerationinherent characteristics of the Apriori algorithmicprogram associated withthe frequent itemset generation andthru a block-based partitioning uses a dynamic work management. The algorithmicprogram greatly enhances the performance and achieves fault tolerancecompared to the presentdistributed Apriori based mostlyapproaches. Planned algorithmic program is enforced and tested over a multinodecluster.

Keywords: Frequent Itemset Mining, Mapreduce, HDFS