A critical review of data leakage detection and prevention techniques: performance assessment and research gaps

Data leakage has become one of the most serious cybersecurity issues in the digital age that resulted in massive financial losses and reputation damage to organizations across the globe. In this paper, the critical review of the current data leakage detection and prevention methods is given to assess the properties of performance and define the critical research gaps. We critically review the use of rule-based methods, machine learning based procedures, hybrid techniques and new technologies such as deep learning and behavioral analytics. What we observed in our evaluation is that, although detection accuracy has advanced considerably, there are very challenging issues to combat in real-time, false positive reduction, and evolutions of attack vectors. The main research gaps found are the lack of performance on insider threats, the lack of processing of encrypted data and lack of scalability to large scale deployment. This review gives researchers and practitioners a clear picture of the capacity of research and the work done and the area of research that will be taken in future.