Wasserstein Generative Adversarial Networks Based Differential Privacy Metaverse Data Sharing.

Area of Science:

• Computer Science
• Cybersecurity
• Artificial Intelligence

Background:

• Differential privacy is crucial for protecting sensitive data in metaverse sharing.
• Random data perturbation can create a trade-off between data utility and privacy.

Purpose of the Study:

• To propose and analyze models and algorithms for differential privacy metaverse data sharing using Wasserstein generative adversarial networks (WGAN).
• To address the imbalance between data utility and privacy in metaverse data sharing.

Main Methods:

• Constructed a mathematical model for differential privacy metaverse data sharing by integrating a regularization term into WGAN.
• Developed and theoretically analyzed basic and federated models and algorithms for WGAN-based differential privacy metaverse data sharing.
• Employed serialized training for the federated model.

Main Results:

• Experimental results validate the theoretical analyses of the proposed algorithms.
• The WGAN-based algorithms demonstrate an effective equilibrium between data privacy and utility.
• Comparative analysis based on utility and privacy metrics confirmed the findings.

Conclusions:

• Wasserstein generative adversarial networks offer a viable solution for secure and useful differential privacy metaverse data sharing.
• The proposed models and algorithms successfully maintain the balance between privacy and utility in metaverse data sharing.
• This research contributes to the advancement of privacy-preserving techniques in virtual environments.