Research on the “Efficiency—Ethics” Tension of Digital Transformation in Educational Evaluation: A Double Helix Mechanism Based on Teacher Data Privacy Protection and Machine Learning Model Optimization

Abstract

The digital transformation of educational evaluation has improved efficiency and accuracy, but it also brings ethical risks between privacy protection and model optimization. The study analyzes the conflict between algorithm interpretability, privacy protection and data transparency, and fi nds that although privacy technology safeguards data security, it may reduce model transparency and increase ethical risks. For this reason, this paper proposes a double helix mechanism to optimize privacy protection and enhance interpretability to promote the educational evaluation system towards intelligence and fairness.