Hadoop Framework to Identify Crop Diseases and Yield Predication

In the growth of Information Technology, Big data come forth as a blazing topic. The main source of human survival depends on agriculture; where it needs a key contribution in the field of crop data analysis. This paper gives a purpose about how to find experiences from accuracy agriculture information through big data approach. In this way, gathering the valuable data in an effective way drives a framework towards major computational challenges in crop analysis where information is remotely gathered. For the storage purpose of huge data availability in agriculture, we are intending Hadoop framework for our work to store a huge volume of crop data. This work gives a better prediction for the farmers to plant which kind of crops to their farm field based on their soil content to improve the productivity. The random forest algorithm is integrated with the MapReduce programming model in Hadoop framework. Data is collected, clean and normalized. The Data is collected from laboratory reports, websites etc. then cleansing of data is done that is main information is extracted from unstructured redundant data. This Normalized data is uploaded on Hadoop Distributed File system and save it in supported form of Hive. HiveQL used to analyze the agricultural data. It is a SQL like query language and by analyzing crop disease symptoms, it finds out disease name and provide a solution based on evidence from historical data. This result is in form of graphs that will helpful for recommending a solution that is high symptoms similarity.