GDPR Compliant Information Confidentiality Preservation in Big Data Processing

Nowadays, new laws and regulations, such as the European General Data Protection Regulation (GDPR), require companies to define privacy policies complying with the preferences of their users. The regulation prescribes expensive penalties for those companies causing the disclosure of sensitive data of their users, even if this occurs accidentally. Thus, it is necessary to devise methods supporting companies in the identification of privacy threats during advanced data manipulation activities. To this end, in this paper, we propose a methodology exploiting relaxed functional dependencies (RFDs) to automatically identify data that could imply the values of sensitive ones, which permits to increase the confidentiality of a dataset while reducing the number of values to be obscured. An experimental evaluation demonstrates the effectiveness of the proposed methodology in increasing compliance to the GDPR data privacy, while reducing the set of values to be partially masked, hence enhancing data usage.