Data obfuscation using a two-layered, controlled approach

Authors

  • Labani Halder
  • Payel Sadhukhan
  • Sarbani Palit Indian Statistical Institute

Keywords:

Application

Abstract

The risks associated with cloud computing increase when data is shared with users who have varying levels of trustworthiness. This work proposes a two-layer data obfuscation protocol to enhance the security in cloud computing when data is shared among users with varying degrees of credibility. The key novelties of the scheme lie in i] its ability to obfuscate the data according to the user's trust and ii] the incorporation of a two-layer mechanism consisting of reversible and partially reversible reconstructions, adding to its robustness. The proposed scheme's computability is validated through experiments on seventeen real-world datasets and supported by a theoretical analysis.

DOI: https://doi.org/10.24135/ICONIP23

Downloads

Published

2025-03-18