Apply an Integrated Responsible AI Framework to Sustain the Assessment of Learning Effectiveness

Tsung-Nan Chou

2023

Abstract

Recent developments in educational data mining and learning analytics have increased the need for explainable artificial intelligence to interpret the decisions or predictions made by the algorithms. In order to analyse the impact of students’ learning input on their learning effectiveness, an innovative responsible and trusted AI framework was developed and implemented as three separate modules that covered five different stages in this study. The first module developed various explainable artificial intelligence (XAI) models based on the model grafting and model fusion techniques that concatenated or synergized a global model with different local models. In addition, the local models were also supplemented by several explanation methods to provide additional explanatory information for the explainable XAI hybrid model. The second module constructed three different safeguard and auditing models to provide complementary predictions for students being misidentified as normal students and discovered the students at risk of failing a course. The adversarial training models developed in the third module applied AI generated synthetic data to train the proposed models and evaluate their performance with an attempt to search for any possible competent models that performed better. The framework was implemented by using Microsoft Power BI tools to create various visualized and interactive dashboards to demonstrate the analysis outcomes.

Paper Citation

in Harvard Style

Chou T. (2023). Apply an Integrated Responsible AI Framework to Sustain the Assessment of Learning Effectiveness. In Proceedings of the 15th International Conference on Computer Supported Education - Volume 2: CSEDU, ISBN 978-989-758-641-5, SciTePress, pages 142-149. DOI: 10.5220/0012058400003470

in Bibtex Style

@conference{csedu23,
author={Tsung-Nan Chou},
title={Apply an Integrated Responsible AI Framework to Sustain the Assessment of Learning Effectiveness},
booktitle={Proceedings of the 15th International Conference on Computer Supported Education - Volume 2: CSEDU,},
year={2023},
pages={142-149},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012058400003470},
isbn={978-989-758-641-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 15th International Conference on Computer Supported Education - Volume 2: CSEDU,
TI - Apply an Integrated Responsible AI Framework to Sustain the Assessment of Learning Effectiveness
SN - 978-989-758-641-5
AU - Chou T.
PY - 2023
SP - 142
EP - 149
DO - 10.5220/0012058400003470
PB - SciTePress