Scholarlink

With the evolving digital forensics to tackle increasingly complex cybercrimes, the volume and sensitivity of data involved in investigations demand innovative approaches to ensure privacy and security. Federated learning (FL), a decentralized machine learning paradigm, has emerged as a promising solution for collaborative forensic analysis without compromising the confidentiality of data Nabavirazavi et al (Lightweight malicious packet classifier for IoT networks check for updates. In Information Security, Privacy and Digital Forensics: Select Proceedings of the International Conference, ICISPD 2022, vol 1075, p 139. Springer Nature, Berlin, 2023); Nabavirazavi and Iyengar (LAKEE: A Lightweight Authenticated Key Exchange Protocol for Power Constrained Devices. arXiv preprint arXiv:2210.16367, 2022). However, the integration of FL in digital forensics introduces unique challenges, particularly in safeguarding the privacy of sensitive evidence while maintaining the integrity and accuracy of forensic models. This chapter delves into the role of privacy-preserving mechanisms in enhancing the applicability of FL for digital forensics. It explores the threats to data privacy in collaborative forensic settings; examines key techniques such as differential privacy, homomorphic encryption, and secure multiparty computation; and highlights their practical applications in forensic investigations. By addressing these aspects, this chapter provides a comprehensive understanding of how privacy-preserving FL can empower forensic analysts to collaborate securely while adhering to strict legal and ethical standards.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Privacy-Preserving AI (Federated Learning) for Digital Forensics

  • 章节

  • 发表日期：

    2025年

查看全文
  • S. S. Iyengar,
  • Seyedsina Nabavirazavi,
  • Yashas Hariprasad,
  • Prasad HB,
  • C. Krishna Mohan

摘要

With the evolving digital forensics to tackle increasingly complex cybercrimes, the volume and sensitivity of data involved in investigations demand innovative approaches to ensure privacy and security. Federated learning (FL), a decentralized machine learning paradigm, has emerged as a promising solution for collaborative forensic analysis without compromising the confidentiality of data Nabavirazavi et al (Lightweight malicious packet classifier for IoT networks check for updates. In Information Security, Privacy and Digital Forensics: Select Proceedings of the International Conference, ICISPD 2022, vol 1075, p 139. Springer Nature, Berlin, 2023); Nabavirazavi and Iyengar (LAKEE: A Lightweight Authenticated Key Exchange Protocol for Power Constrained Devices. arXiv preprint arXiv:2210.16367, 2022). However, the integration of FL in digital forensics introduces unique challenges, particularly in safeguarding the privacy of sensitive evidence while maintaining the integrity and accuracy of forensic models. This chapter delves into the role of privacy-preserving mechanisms in enhancing the applicability of FL for digital forensics. It explores the threats to data privacy in collaborative forensic settings; examines key techniques such as differential privacy, homomorphic encryption, and secure multiparty computation; and highlights their practical applications in forensic investigations. By addressing these aspects, this chapter provides a comprehensive understanding of how privacy-preserving FL can empower forensic analysts to collaborate securely while adhering to strict legal and ethical standards.