| Autor: |
Shibani, Antonette, Ratnavel Rajalakshmi, Srivarshan Selvaraj, Faerie Mattins, Dhivya Chinnappa |
| Přispěvatelé: |
Mingyu Feng, Tanja Käser, Partha Talukdar |
| Rok vydání: |
2023 |
| DOI: |
10.5281/zenodo.8115750 |
| Popis: |
Recent works in educational data mining emphasize the need for producing practical insights that enhance learning. There is particular interest in supporting student writing by generating actionable feedback using machine learning algorithms. While algorithmic efficiency is generally sought after in machine learning, it might not be the most important aspect to consider for ``explainability''. This paper presents a predictive model for argumentative writing feedback where explanations supported by Local Interpretable Model-agnostic Explanations (LIME), SHapley Additive explanation (SHAP), and logic are derived to generate insights for designing student feedback on argumentative writing. It discusses the computational trade-offs and insights derived that inform writing feedback in practice, with lessons transferable to other contexts. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
