Single Label Model for Confidentiality in Big Data

Abstract

Data security is defined as preventing actions such as accessing, using, changing, and eliminating information without authorization and consists of specific basic elements called confidentiality, integrity and accessibility. Many studies have been carried out in the literature to ensure information security. However, no study has been carried out to provide these basic elements together. In this study, we created a model that provides three basic elements of information security for large data. With the single labeling model we propose, a more practical and flexible structure has been realized for all operations performed in the database (reading, writing, updating, deleting) on actual data. In previous studies with the labeling model, a separate label was used for only reading or writing only and a structure that will ensure both confidentiality and integrity at the same time has not been established. In our study, we have shown how to perform an authorization and access control between which operation and which user by looking at a single label for all operations performed on the data. In this way, unlike the studies in the literature, data confidentiality, data integrity and data consistency were provided for all processes. The results of the proposed single-label model have been shown in comparison with the application and experimental study we have carried out. The results obtained are promising for future studies.