EUVIMEDEUVIMED entdecken
Crossref

From Student-Risk Prediction to Educational Action: Integrating Explainability, Human Oversight, and Intervention Evaluation in Higher Education

Nandan Limakrisna

International Journal of Artificial Intelligence, Machine Learning and Data Analytics · 26.10.2026 · 1 (2) · 12-18

DOI: 10.68033/ijaimda.18 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Artificial Intelligence, Machine Learning and Data Analytics, online erschienen am 26.10.2026. Verfasst von Nandan Limakrisna. Die Arbeit nennt 10 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Predictive learning analytics can help higher education institutions identify students who may need academic support, but the usefulness of a risk prediction depends on what happens after the prediction is generated. This paper develops a critical integrative review and conceptual framework connecting explainability, human oversight, intervention pathways, and evaluation in educational decision support. The review draws on selected peer-reviewed review articles, research publications, and policy guidance concerning higher-education learning analytics, student-risk prediction, explainable artificial intelligence (XAI), and responsible governance. The synthesis is organised around four questions: what is predicted; how model outputs and uncertainty are explained; how staff translate alerts into supportive action; and how technical and educational outcomes are evaluated. The paper proposes a decision-to-intervention framework linking data suitability, predictive modelling, explanation and reliability checks, human review, proportionate student support, and outcome evaluation. Privacy, fairness, accountability, and contestability are treated as lifecycle requirements. The framework's contribution is to connect prediction quality to the institutional pathway through which alerts are reviewed, acted upon, and evaluated. The paper is an integrative review and conceptual contribution rather than a completed systematic literature review; it does not claim exhaustive database coverage, PRISMA screening counts, or pooled effect estimates. Future empirical work should validate the framework across institutions and assess whether the decision-to-intervention pathway improves support quality, equity, and educational outcomes.

Zitieren

APA

Limakrisna, N. (2026). From Student-Risk Prediction to Educational Action: Integrating Explainability, Human Oversight, and Intervention Evaluation in Higher Education. International Journal of Artificial Intelligence, Machine Learning and Data Analytics, 1(2), 12-18. https://doi.org/10.68033/ijaimda.18

Vancouver

Limakrisna N. From Student-Risk Prediction to Educational Action: Integrating Explainability, Human Oversight, and Intervention Evaluation in Higher Education. International Journal of Artificial Intelligence, Machine Learning and Data Analytics. 2026;1(2):12-18. doi:10.68033/ijaimda.18

BibTeX
@article{Limakrisna2026,
  author = {Limakrisna, Nandan},
  title = {From Student-Risk Prediction to Educational Action: Integrating Explainability, Human Oversight, and Intervention Evaluation in Higher Education},
  journal = {International Journal of Artificial Intelligence, Machine Learning and Data Analytics},
  year = {2026},
  volume = {1},
  number = {2},
  pages = {12--18},
  doi = {10.68033/ijaimda.18}
}
Kostenlos starten →