Scholarlink

<p>In mobile educational networks, where data is shared between devices and cloud platforms, AI-based security and privacy advancements protect critical student data. Mobile learning applications are growing rapidly, making security and personalization a necessity. Existing techniques frequently use centralized data storage and processing, which promotes data breaches, illegal access, and privacy violations. In dynamic mobile contexts, standard encryption cannot completely defend against inference attacks or preserve real-time performance. The proposed framework integrates Federated Learning with Differential Privacy (FL-DP) to address these challenges. In this approach, student data remains on their devices, and only encrypted, noise-added model updates are transmitted to a central server for aggregation. This allows collaborative model training without revealing raw data while protecting privacy. This technique enables a customized learning recommendation system that adapts educational material to students’ learning behavior while complying with privacy laws. Its 96.2% privacy protection, 3.1% data leakage reduction, and 97.8% model accuracy make FL-DP a solid solution for safe mobile educational environments.</p>

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

AI-based security and privacy enhancements for student data protection in mobile educational networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月31日

查看全文
  • Junfeng Qiang,
  • Yun Liu

摘要

In mobile educational networks, where data is shared between devices and cloud platforms, AI-based security and privacy advancements protect critical student data. Mobile learning applications are growing rapidly, making security and personalization a necessity. Existing techniques frequently use centralized data storage and processing, which promotes data breaches, illegal access, and privacy violations. In dynamic mobile contexts, standard encryption cannot completely defend against inference attacks or preserve real-time performance. The proposed framework integrates Federated Learning with Differential Privacy (FL-DP) to address these challenges. In this approach, student data remains on their devices, and only encrypted, noise-added model updates are transmitted to a central server for aggregation. This allows collaborative model training without revealing raw data while protecting privacy. This technique enables a customized learning recommendation system that adapts educational material to students’ learning behavior while complying with privacy laws. Its 96.2% privacy protection, 3.1% data leakage reduction, and 97.8% model accuracy make FL-DP a solid solution for safe mobile educational environments.