A Robust Data Anonymization Approach Based on Enhanced Starfish Optimization and Levy Flight for Social Networks

• Published in: SN computer science Vol. 6; no. 7; p. 778
• Main Authors: Sivasankari, K., Uma Maheswari, K. M.
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.10.2025; Springer Nature B.V
• Subjects: Algorithms; Big Data; Clustering; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Confidentiality; Data analysis; Data collection; Data Structures and Information Theory; Datasets; Identity; Information retrieval; Information Systems and Communication Service; Methods; Optimization; Original Research; Pattern Recognition and Graphics; Performance enhancement; Performance evaluation; Performance measurement; Privacy; Publishing; Social networks; Software Engineering/Programming and Operating Systems; Vision; Samoa; K-anonymization; Levy flight mechanism; Enhanced starfish optimization algorithm; Information loss; Degree of anonymization
• ISSN: 2662-995X; 2661-8907
• DOI: 10.1007/s42979-025-04285-7

With the rapid growth of data collection in various sectors, safeguarding individual privacy while enabling meaningful data analysis has become a significant challenge. Data anonymization plays a vital role in ensuring the privacy of individuals when sharing datasets for research, policy-making, or commercial use. However, traditional anonymization techniques often suffer from high information loss and insufficient privacy guarantees, especially when applied to large or complex datasets. Moreover, many optimization-based anonymization approaches are prone to getting trapped in local optima, which reduces the quality and effectiveness of anonymization. To address these challenges, this research proposes a novel privacy-preserving framework that integrates the Enhanced Starfish Optimization Algorithm (ESOA) with Levy Flight to improve the performance of data anonymization. The approach begins with a rigorous data pre-processing phase to clean and normalize the input dataset, ensuring consistency and reliability. The ESOA is then employed to perform anonymization, optimizing the trade-off between data utility and privacy. The integration of the Levy Flight mechanism enhances the exploratory capabilities of the optimization process, helping the algorithm avoid local minima and achieve better anonymization results. The proposed method is evaluated using critical performance metrics such as Information Loss, Degree of Anonymization (DA), and execution time. Experimental results indicate that the method effectively minimizes information loss while maximizing privacy protection and computational efficiency. This makes the approach a promising solution for secure and efficient data publishing in various privacy-sensitive applications.