Optimal Data Privacy Protection: K-Anonymity Approach
Suttinee Sawadsitang, Waranya Mahanan, Dusit Tao Niyato · 2024
Digital advancements have facilitated vast data storage and sharing capabilities, yet ease of access often neglects data sensitivity, leading to privacy risks and leaks. This is especially critical in healthcare, where staff may inadvertently expose patient information. Such lapses heighten the risk of exploitation, exemplified by widespread phishing and scams targeting personal data. To mitigate these risks, this paper introduces the Optimal Data Privacy Protection (ODPP) system. We formulation a mix-integer programming problem subject to k-anonymity approach to generalize numerical data. The ODPP system helps a novice user automates data anonymization while minimizing total degree of information loss. We evaluate the system effectiveness with (i) sample data and (ii) healthcare dataset on stroke, hypertension, and heart disease, highlighting its potential to bolster data privacy across domains.