Enhancement of Classification using FPFF-ANN for Big data Analysis in Distributed Environment

Abstract

The development of massive amount of information from any source of group at any time, wherever and from any device which is termed as Big Data. The age group of big data becomes a dangerous challenge to grip, take out and access these data is short length of time. The detection of everyday itemsets is an significant issue of data mining which helps in engendering the qualitative information for the business insight and helps for the verdict makers. For the extracting the necessary itemsets from the big data a variety of big data logical techniques has been evolved such as relationship rule mining, genetic algorithm, mechanism learning, FP-growth algorithm etc. In this paper we suggest FP-ANN algorithm to promote the FP enlargement calculation with neural networks to maintain the feed forward approach. The recommend algorithm uses the Twitter social dataset for the collection of frequent itemsets and the proportional analysis of this approach is done using the different performance measuring parameters such as Precision, Recall, F-measure, Time complexity, Computation cost and time. The simulation of proposed work is done using the JDK, JavaBeans, and Wamp server software. The experimental results of projected algorithm gives better results in deference of time difficulty, computation cost and time also. It also gives enhanced results for the Precision, recall and F-measure.