Distortion Techniques for Profile Matching Attacks in Online Social Networks

This repository contains our research on various distortion techniques to mitigate the risk of profile matching attacks in online social networks (OSNs).

📄 Paper

• Download PDF

🔬 Overview

• Objective: Compare and analyze distortion techniques to prevent profile matching attacks.
• Techniques Studied:
  • Data Perturbation
  • Noise Addition
  • Anonymization
  • Tokenization
  • Hashing
  • Suppression
  • Generalization

📊 Key Findings

• Anonymization had the highest robustness in preventing profile matching attacks.
• Noise addition, while useful, had lower robustness due to potential reversibility.
• Hybrid approaches combining multiple techniques showed promise in increasing privacy without sacrificing utility.

🖥️ Code & Datasets

If we release any datasets or code used for analysis, they will be available in the /code/ folder.

🏷 Citation

If you use this work, please cite it as follows:

@misc{dellal2024distortion,
  title={Comparison and Analysis of Distortion Techniques in Terms of Mitigating the Risk of Profile Matching Attacks in Online Social Networks},
  author={Zeynep Doğa Dellal, Borga Haktan Bilen, Alper Bozkurt, İzgi Nur Tamcı, Gizem Gökçe Işık},
  year={2024},
  url={https://github.com/zedyjy/Distortion-Techniques-Profile-Matching}
}