Context-Aware Attentive Knowledge Tracing

TL;DR

This paper introduces AKT, an interpretable neural network model for knowledge tracing that uses attention mechanisms and psychometric regularization to improve prediction accuracy and provide actionable feedback for personalized learning.

Contribution

AKT combines attention-based neural networks with cognitive-inspired components, including a novel monotonic attention mechanism and Rasch model regularization, enhancing interpretability and performance in knowledge tracing.

Findings

AKT outperforms existing methods by up to 6% in AUC on benchmark datasets.

The model provides interpretable feedback suitable for personalized education.

AKT demonstrates strong potential for automated feedback and learner personalization.

Abstract

Knowledge tracing (KT) refers to the problem of predicting future learner performance given their past performance in educational applications. Recent developments in KT using flexible deep neural network-based models excel at this task. However, these models often offer limited interpretability, thus making them insufficient for personalized learning, which requires using interpretable feedback and actionable recommendations to help learners achieve better learning outcomes. In this paper, we propose attentive knowledge tracing (AKT), which couples flexible attention-based neural network models with a series of novel, interpretable model components inspired by cognitive and psychometric models. AKT uses a novel monotonic attention mechanism that relates a learner's future responses to assessment questions to their past responses; attention weights are computed using exponential decay and a context-aware relative distance measure, in addition to the similarity between questions. Moreover, we use the Rasch model to regularize the concept and question embeddings; these embeddings are able to capture individual differences among questions on the same concept without using an excessive number of parameters. We conduct experiments on several real-world benchmark datasets and show that AKT outperforms existing KT methods (by up to $6\%$ in AUC in some cases) on predicting future learner responses. We also conduct several case studies and show that AKT exhibits excellent interpretability and thus has potential for automated feedback and personalization in real-world educational settings.