Prediction-Based Parallel Clustering Algorithm for M-Commerce

Abstract

A rapid increase of mobile commerce (M-Commerce) with sensing devices has resulted in enormous services from different service providers. M-Commerce services provide numerous ranges of emerging services. Also, different qualitative matrices are provided with similar functionality. Automatically the service flow is combined with other services. M-Commerce stakeholders are ambient, dynamic in nature, which requires efficient techniques to enhance the output. Major challenge is to select appropriate optimization technique or algorithm to provide a numeric set of services with dynamic qualities. It is difficult to propose a method directly to predict M-Commerce. Hence, this research proposes a method of prediction in M-Commerce techniques proposed a prediction based parallel clustering algorithm using hybrid optimization technique for M-Commerce. Hybrid optimization technique or algorithm can be developed by applying cross-mutation technique in adaptive ant colony optimization with particle swarm optimization to improve the efficiency and throughput of the system in M-Commerce. To predict the optimum service it runs in parallel using MapReduce on a Hadoop platform. Parallel processing services reduce the time factor, which is essential for processing the massive amount of unstructured data in a mobile environment. Relevancy, correctability of this proposed system would be validated through simulation and modeling on real-time existing data sets.