Article English OPEN
Florin Ogigau-Neamtiu (2017)
  • Publisher: Regional Department of Defense Resources Management Studies
  • Journal: Journal of Defense Resources Management (issn: 2068-9403, eissn: 2068-9403)
  • Subject: data | obfuscation | encryption | automatization | machine learning | natural language processing | Military Science | U

Contemporary organizations face big data security challenges in the cyber environment due to modern threats and actual business working model which relies heavily on collaboration, data sharing, tool integration, increased mobility, etc. The nowadays data classification and data obfuscation selection processes (encryption, masking or tokenization) suffer because of the human implication in the process. Organizations need to shirk data security domain by classifying information based on its importance, conduct risk assessment plans and use the most cost effective data obfuscation technique. The paper proposes a new model for data protection by using automated machine decision making procedures to classify data and to select the appropriate data obfuscation technique. The proposed system uses natural language processing capabilities to analyze input data and to select the best course of action. The system has capabilities to learn from previous experiences thus improving itself and reducing the risk of wrong data classification.
Share - Bookmark