Misconception Diagnosis From Student-Tutor Dialogue: Generate, Retrieve, Rerank

TL;DR

This paper introduces a novel LLM-based method for detecting student misconceptions in dialogues by generating, retrieving, and reranking candidate misconceptions, significantly improving accuracy over baseline models.

Contribution

The work presents a new multi-step approach using fine-tuned LLMs for misconception detection, combining generation, retrieval, and reranking to enhance performance.

Findings

Improved misconception detection accuracy over baseline models

Fine-tuning enhances both generation quality and model performance

Ablation studies validate the importance of generation and reranking steps

Abstract

Timely and accurate identification of student misconceptions is key to improving learning outcomes and pre-empting the compounding of student errors. However, this task is highly dependent on the effort and intuition of the teacher. In this work, we present a novel approach for detecting misconceptions from student-tutor dialogues using large language models (LLMs). First, we use a fine-tuned LLM to generate plausible misconceptions, and then retrieve the most promising candidates among these using embedding similarity with the input dialogue. These candidates are then assessed and re-ranked by another fine-tuned LLM to improve misconception relevance. Empirically, we evaluate our system on real dialogues from an educational tutoring platform. We consider multiple base LLM models including LLaMA, Qwen and Claude on zero-shot and fine-tuned settings. We find that our approach improves predictive performance over baseline models and that fine-tuning improves both generated misconception quality and can outperform larger closed-source models. Finally, we conduct ablation studies to both validate the importance of our generation and reranking steps on misconception generation quality.