Privacy and security has become major issues in today's communication. In respect to the last ten years the nature and utilization of the communication technology is changed much frequently. Due to this a significant amount of data is communicated in a fraction of seconds. Therefore the traditional computational techniques have moved towards the big data processing and analytics. In this environment the entire client module and administration is directly connected with the same data sources. Due to this the communication becomes easy but security and privacy concern in communicated data has appeared. Sometimes these issues are arising due to data leakage. In this presented work the main aim is to investigate the privacy and security concerns due to data leakage in big data environment. The main reason to utilize the big data is to demonstrate the real time system using twitter accounts to fetch and improve the sensitivity of data. During the investigation the promising approach is appeared where the sliding window and fuzzy logic based system is provided to analyze and reform the data. But this approach is found slow processing capability by which the system performance is affected. Due to this a new approach using the random walk technique is prepared by modification of existing system to enhance the resource consumption of the system.

Data integrity and storage efficiency are two essential needs for cloud storage. Cloud computing has emerged as a long-dreamt vision of the utility computing paradigm that provides reliable and resilient infrastructure for users to remotely store data and use on-demand applications and services. Currently, many individuals and organizations mitigate the burden of local data storage and reduce the maintenance cost by outsourcing data to the cloud. However, the outsourced data is not always trustworthy due to the loss of physical control and possession over the data. With the growing awareness of data privacy, more and more cloud users choose to encrypt their sensitive data before outsourcing them to the cloud. Therefore, in order to resolve the issues in cryptographic security a new methodology is required to develop. That provides the data owner and data management with efficient. Thus, the proposed solution incorporates the TF-IDF for calculating frequency of each word of text data and performing indexing on selected words. In this paper, the investigation is done about the outsourced data and their sensitivity and security issues. In this, a mechanism is proposed for public cloud data security by means of BST-MRPAS i.e. Binary Search Tree based Multi Replica for Public Cloud Auditing System for the end user applications. Additionally for providing end user trust and security management the upload, download and Update services are provided.