Hongda Jiang

Hongda Jiang

August 23, 2026

AI-enabled student support, Student profiling, Actionable labels, Learning analytics ethics, Digital stereotyping.

AI-enabled student-support systems increasingly translate digital traces into profiles, risk estimates, and recommendations for educators. This study asks how a provisional description becomes actionable in Chinese universities. Through conceptual reconstruction, normative interpretation, and structured mapping of official documents from four universities, it develops a five-transition framework: data proxying, profile labeling, probabilistic inference, educational intervention, and feedback. Public documents identify functional entry points but do not verify effects or continuous mechanisms. The paper argues that digital stereotyping arises only when a purpose, evidence, action, or time boundary is crossed, institutional treatment changes, and correction mechanisms fail. It then specifies input, decision, action, and exit controls that keep high-impact transitions explainable, contestable, suspendable, and correctable. The framework connects human-centred learning analytics with the institutional use of student profiles while explicitly limiting empirical claims.

© 2026, the Authors. Published by ISCCAC

This is an open access article distributed under the CC BY-NC license