A Conceptual Model for End-to-End Causal Discovery in Knowledge Tracing

TL;DR

This paper introduces a novel causal discovery approach in knowledge tracing using a causal gated recurrent unit, aiming to uncover skill relationships from student response data to improve educational strategies.

Contribution

It proposes a new causal GRU module with learnable matrices for causal ordering and structure, enabling end-to-end differentiable learning in knowledge tracing models.

Findings

Placed among top entries in NeurIPS 2022 Challenge

Demonstrated preliminary effectiveness on public leaderboard

Proposed a novel causal modeling component for knowledge tracing

Abstract

In this paper, we take a preliminary step towards solving the problem of causal discovery in knowledge tracing, i.e., finding the underlying causal relationship among different skills from real-world student response data. This problem is important since it can potentially help us understand the causal relationship between different skills without extensive A/B testing, which can potentially help educators to design better curricula according to skill prerequisite information. Specifically, we propose a conceptual solution, a novel causal gated recurrent unit (GRU) module in a modified deep knowledge tracing model, which uses i) a learnable permutation matrix for causal ordering among skills and ii) an optionally learnable lower-triangular matrix for causal structure among skills. We also detail how to learn the model parameters in an end-to-end, differentiable way. Our solution placed among the top entries in Task 3 of the NeurIPS 2022 Challenge on Causal Insights for Learning Paths in Education. We detail preliminary experiments as evaluated on the challenge's public leaderboard since the ground truth causal structure has not been publicly released, making detailed local evaluation impossible.