Jan Zizka et al. (Eds) : ICAITA, SAI, CDKP, Signal, NCO - 2015 pp. 183–198, 2015. © CS & IT-CSCP 2015 DOI : 10.5121/csit.2015.51515               Amine RAHMANI, Abdelmalek AMINE and Mohamed Reda HAMOU GeCoDe Laboratory, Department of Informatics, Dr. TAHAR Moulay university of Saida – Algeria- Aminerahmani2091@gmail.com, amine_abd1@yahoo.fr, hamoureda@yahoo.fr ABSTRACT Nowadays, the concept of big data grows incessantly; recent researches proved that 90% of the whole data existed on the web had been created in last two years. However, this growing bumped by many critical challenges resides generally in security level; the users care about how could providers protect their privacy on their data. Access control, cryptography, and de- identification are the main search areas grouped under a specific domain known as Privacy Preserving Data Publishing. In this paper, we bring in suggestion a new model for access control over big data using digital signature and confidence interval; we first introduce our work by presenting some general concepts used to build our approach then presenting the idea of this report and finally we evaluate our system by conducting several experiments and showing and discussing the results that we got. KEYWORDS Access control, standard deviation, privacy preserving, big data, numeric signature, confidence interval 1. INTRODUCTION Privacy, timeless, scalability of data is the most important problems that big data recognize starting from the first step of data acquisition; in fact, one of the most disturbed principle that are used in big data is the fact of losing control on data. This concept led to a lot of criticism from clients, losing control on your own data means losing everything related to the control even the access control. Before the coming of the concept big data, controlling access on such data was done locally using the known models such as mandatory models (MAC), discriminatory models (DAC) or role based models (RBAC) but those last cannot be used because of some impediments; in case of DAC models the users defines the right access by himself while in the use of big data the user lose the entire control on his data; in case of MAC models the right access are defined by a major