International Journal of Engineering and Technology

the big data is difficult to be analyzed due to the presence and characteristics of huge amount of data. Hadoop technology plays a key role in analyzing the large scale data. The aggregate queries are executed on more columns concurrently and it is difficult for huge amount of data. This paper is proposing the method in which the fast RAQ is dividing the big data in to autonomous partitions by means of a balanced partition algorithm and later for each partition a local assessment sketch is generated. By the arrival of the range-aggregate query demand the fast RAQ gets the result in a direct manner by shortening local estimate from all partition and then the cooperative results are provided. Thus in fast RAQ technique three tier Architecture is insisted and they are of 1.Extracting the helpful information’s from Unstructured Data, 2.Implementation of the big data in Multi system Approach, 3.Application Deployment – Insurance/ Banking. This paper is implement for the banking domain process and two major departments are involved in this process and they are 1.To maintain the accounts and for adding new clients the Bank Server is used. To create account in any bank the user have to give their ID proof at the time of registration.2.Account Monitoring Server is used for monitoring every users accounts in various banks and this server is used for retrieving the users who are maintaining and transacting more than Rs 50,000 per annum in various bank accounts by using the similar ID proof is identified by Map Reduce technique. The Online Aggregation is a smart sampling-based method that is performed to provide response to aggregation query by an approximation to the last outcome, with the self-assurance interval which is becoming tighter eventually. It is built into a Map-Reduce-based cloud scheme for analytics of the big data that allows the user to save the money by means of killing the calculation early and to observe the query progress when the enough accuracy is achieved.