Fast and efficient agro data classification model for agriculture management system using hierarchical cloud computing

Abstract

Data analytics (DA), Internet of Things (IoT) and cloud computing framework are employed to build a cost efficient and productive agriculture management system. The remote sensing forecasting and GIS Technology provide various sensory information to stake holders/users such as rainfall pattern, weather related data (such as temperature, humidity, pressure etc.). These sensory data are of unstructured format. The existing system lack efficiency in performing analysis on such data. Since it fails to bring good tradeoff between speedup and memory efficiency. To overcome these research challenges, this work presents an Accurate Classification Model (ACM) for Agriculture Management System (AMS). Firstly, a selective clustering algorithm is proposed to classify unstructured multidimensional selective agriculture data to structured format. Further, this work presents a novel hierarchical clustering model to perform clustering on output data of selective clustering algorithm and stores the data on standard Hierarchical cloud storage architecture. A parallel algorithm to perform classification of structured data using Hadoop MapReduce framework is presented. Experiments are conducted on real-time agricultural data. The results obtained indicate a considerable improvement over exiting model in terms of computation cost, latency, accuracy, memory efficiency and speedup.