Adaptive-parameter memetic algorithm for privacy-preserving trajectory data publishing: A multi-objective optimization approach

Abstract Trajectory data has grown pervasive, benefiting practical applications, including transportation administration and location-based operations. Nevertheless, trajectories may reveal extremely sensitive information about an individual, including movement patterns, personal profiles, geographical locations, and social contacts, necessitating privacy protection while disseminating trajectory data. Therefore, prioritizing privacy protection is crucial while analyzing trajectory data. Current methods of protecting privacy concentrate on single objective optimizing techniques such as maximizing data utility but often disregard various privacy constraints. To overcome this challenge, this study aims to improve both data privacy and usability by balancing competing goals-maximizing privacy while maintaining useful information-through a Multi-Objective Optimization (MOO) approach in trajectory data publishing. We provide a unique algorithm named Adaptive-Parameter Memetic Algorithm (APMA) that employs a non-dominated sorting multi-objective technique and a Memetic Algorithm (MA). This algorithm utilizes adaptive memory-based mutation and crossover strategies to dynamically adjust the mutation and crossover parameters and improve the solution’s quality. The proposed innovative local search strategy helps to achieve better population diversity and solution quality. Comprehensive studies illustrate the efficacy of the proposed method regarding solution quality and convergence outcomes.