Automatic Assessment of Students' Classroom Engagement with Bias Mitigated Multi-task Model

2025-10-24

AI Summary

To address bias and poor interpretability in automated assessment of student classroom engagement in online education, this paper proposes a fairness-aware and interpretable multi-task learning framework. Methodologically, it integrates transfer learning with attribute-orthogonal regularization to explicitly disentangle sensitive attributes (e.g., gender) in the segmentation model’s classifier, thereby suppressing reliance on such features; concurrently, a multi-task architecture jointly optimizes engagement prediction and learning of attribute-invariant representations. Experiments demonstrate that the approach increases the Pearson correlation coefficient between prediction distributions across sensitive groups from 0.897 to 0.999, substantially mitigating group-level bias. The resulting model achieves both high predictive accuracy and ethical compliance. The implementation is publicly available.

Technology Category

• Machine Learning: Ethics, Bias, and Fairness
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

With the rise of online and virtual learning, monitoring and enhancing student engagement have become an important aspect of effective education. Traditional methods of assessing a student's involvement might not be applicable directly to virtual environments. In this study, we focused on this problem and addressed the need to develop an automated system to detect student engagement levels during online learning. We proposed a novel training method which can discourage a model from leveraging sensitive features like gender for its predictions. The proposed method offers benefits not only in the enforcement of ethical standards, but also to enhance interpretability of the model predictions. We applied an attribute-orthogonal regularization technique to a split-model classifier, which uses multiple transfer learning strategies to achieve effective results in reducing disparity in the distribution of prediction for sensitivity groups from a Pearson correlation coefficient of 0.897 for the unmitigated model, to 0.999 for the mitigated model. The source code for this project is available on https://github.com/ashiskb/elearning-engagement-study .

Problem

• Automatically assessing student engagement in online learning environments
• Mitigating bias from sensitive attributes like gender in predictions
• Developing interpretable multi-task models for educational applications

Innovation

• Attribute-orthogonal regularization mitigates bias in engagement assessment
• Split-model classifier uses multi-task transfer learning strategies
• Reduces prediction disparity across sensitive groups effectively