MapReduce Framework Based Sequential Association Rule Mining with Deep Learning Enabled Classification in Retail Scenario

Abstract

Association Rule Mining is used for data mining using the associate rules for detecting frequent items. Hence, this research introduces a novel sequential rule mining for predicting user category using the proposed hybrid Tasmanian Water Devil Optimization based map reduce framework. The introduced Tasmanian Water Devil Optimization integrates the feeding behavior of the Tasmanian devil in the Tasmanian devil optimization with the water cycle process of streams and rivers flowing toward the sea in the water cycle algorithm for choosing the optimal rule from the rule set. In this, the big data based on the retail transaction is utilized for identifying the frequent items using the proposed Tasmanian Water Devil Optimization Sequential Associate Mining technique at the Mapper side. Then, the integration of rules is employed at the Reducer side for the prediction of user category using the Deep Convolutional Neural Network. Thus, the proposed Tasmanian Water Devil Optimization Sequential Associate Mining + Deep Convolutional Neural Network outperformed other conventional methods, obtained enhanced performance in terms of rules mined, execution time, recall, and precision, and obtained the values of 183, 7.70, 0.88, and 0.88, respectively.