INPUT SPLIT FREQUENT PATTERN TREE USING MAPREDUCE PARADIGM IN HADOOP

Big data has been attracted in information industry and in the society in the recent years, due to the wide availability of huge amount of data in the Internet and the complexity of data is growing every day. Hence, distributed data mining algorithms has decided to exploit big data adaptable to current technology. Since there exist some limitations in traditional algorithm for dealing with the massive volume of data set which degrades the performance. So, thereby the authors require fast and efficient scalable frequent item sets for storing and processing large data sets. Existing algorithm like apriori algorithm performs a multiple scans from external storage, which leads to heavy burden for I/O devices. In this article, the authors proposed Association Rule Mining based on Hadoop Distributed File System for storing huge amount of data and implemented using MapReduce object oriented programming paradigm for processing of a data.