Date of Publication :10th August 2017

Abstract: Now a days Big Data has defined very large amount of data, it includes both structured and unstructured format. The structured data analyzing is very easy task but an unstructured data analyzing is very difficult that can be produced by an individuals (eg. Twitter data)it also gathered by sensors(eg. satellites, videos) which can range from giga bytes, tera bytes and peta bytes. Big Data entitles more and more data that can be analyzed through various analyzing techniques. If the right analytic method is applied to unstructured datasets we can easily analyze and classifying various patterns, But at the same time will consider efficiency and scale of Data. In the real world the major issue of Big Data is early warning predictions is the use of Satellite imagery and Radar Sensor data. In the Satellite imagery data could reach a million derived spatial objects such data querying ,managing and various image patterns classification is very difficult task. So a proper architecture should be proposed to gain knowledge about Big Data for analyzing various Satellite imagery patterns classifications with hadoop technology. In the proposed architecture differentiate various classification methods for various satellite imagery pattern classification methods and also proposing Google’s Map reduce C4.5 Algorithm for effective classification to increase performance of patterns classification and increasingly large volume of Data sets to results both time efficiency and scalabilty. This research is carrying based on NASA Satellite data and Twitter data and also in weather forecasting.

Reference :