Exploring Fairness in Educational Data Mining in the Context of the Right to be Forgotten

TL;DR

A novel class of selective forgetting attacks designed to compromise the fairness of learning models while maintaining their predictive accuracy, thereby preventing the model owner from detecting the degradation in model performance is introduced.

Abstract

In education data mining (EDM) communities, machine learning has achieved remarkable success in discovering patterns and structures to tackle educational challenges. Notably, fairness and algorithmic bias have gained attention in learning analytics of EDM. With the increasing demand for the right to be forgotten, there is a growing need for machine learning models to forget sensitive data and its impact, particularly within the realm of EDM. The paradigm of selective forgetting, also known as machine unlearning, has been extensively studied to address this need by eliminating the influence of specific data from a pre-trained model without complete retraining. However, existing research assumes that interactive data removal operations are conducted in secure and reliable environments, neglecting potential malicious unlearning requests to undermine the fairness of machine learning systems. In this paper, we introduce a novel class of selective forgetting attacks designed to compromise the fairness of learning models while maintaining their predictive accuracy, thereby preventing the model owner from detecting the degradation in model performance. Additionally, we propose an innovative optimization framework for selective forgetting attacks, capable of generating malicious unlearning requests across various attack scenarios. We validate the effectiveness of our proposed selective forgetting attacks on fairness through extensive experiments using diverse EDM datasets.

Figures (5)

• Figure 1: Overview of selective forgetting attacks in educational data mining systems. The attacker aims to make malicious unlearning requests to the model owner. Upon completion of the unlearning process, the resulting unlearned model exhibits biases to inputs, exacerbating the fairness gap.
• Figure 2: AEOD increment ratio for whole unlearning on OULAD, Student Performance, and xAPI-Edu-Data.
• Figure 3: AEOD increment ratio for partial unlearning on OULAD, Student Performance, and xAPI-Edu-Data.
• Figure 4: (a) AEOD increment ratio for different unlearning methods on Student Performance. (b) AEOD increment ratio for different sensitive features on OULAD. (c) AEOD increment ratio for different fairness losses on xAPI-Edu-Data.
• Figure 5: AEOD increment ratio of selective forgetting attacks in the black-box setting.