Title: Sensitive data hiding in financial anti-fraud process

Authors: Vassilios S. Verykios; Elias C. Stavropoulos; Vasilis Zorkadis; George Katsikatsos; Evangelos Sakkopoulos

Addresses: School of Science and Technology, Hellenic Open University, 18 Par. Aristotelous str, 26335, Patras, Greece ' School of Science and Technology, Hellenic Open University, 18 Par. Aristotelous str, 26335, Patras, Greece ' Hellenic Data Protection Authority, Kifissias 1-3, 115 23, Athens, Greece ' Department of Informatics, University of Piraeus, Karaoli and Dimitriou 80, Piraeus, Greece ' Department of Informatics, University of Piraeus, Karaoli and Dimitriou 80, Piraeus, Greece

Abstract: This research presents an approach to protect personally identifiable information in compliance with the national and European institutional data protection framework in a way that still allows interoperability of information systems and applications. It is proposed to adopt privacy-preserving information hiding techniques to facilitate targeted data mining without infringing privacy restrictions. This approach is proposed as a strategic tool in the fight against financial and insurance fraud. To resolve issues related to the implementation of the protected registration interface process, the research team is turning attention to the development of algorithms and approaches based on intelligent itemset hiding. The research proposal attempts to contribute to the strategic modernisation of public authorities and financial organisations, aiming at the production of original software to provide services to them, facilitating and accelerating the work to combat fraud. The approach is analytically prevailing on previous approaches and it has experimentally shown encouraging results.

Keywords: data hiding; privacy preserving data mining; knowledge hiding; frequent itemset hiding; constraint-based data mining; anti-fraud; sensitive data.

DOI: 10.1504/IJEG.2022.123253

International Journal of Electronic Governance, 2022 Vol.14 No.1/2, pp.7 - 27

Received: 12 Oct 2020
Accepted: 09 Sep 2021
Published online: 06 Jun 2022 *