PP-TEBDE: Privacy-Preserving Temporal Exposure Bias Detection and Explanation

PP-TEBDE is a novel method for detecting and mitigating temporal exposure bias in recommender systems while preserving user privacy and providing explainable insights.

Features

• Temporal exposure bias detection using attention mechanisms and causal inference
• Privacy-preserving framework combining federated learning, homomorphic encryption, and differential privacy
• Explainable bias insights generation
• Adaptive debiasing for dynamic recommendation adjustment
• Application to educational recommender systems