用课堂对话数据提升数学辅导中的对话分析模型表现
Enhancing Talk Moves Analysis in Mathematics Tutoring through Classroom Teaching Discourse
- 利用课堂对话预训练增强辅导场景下的对话行为识别
- 加入更长上下文与说话人信息后，模型准确率显著提升
- 适合教育人工智能、对话系统研究者参考
人类辅导在支持学生学习、提升学业表现和促进个人成长方面至关重要。本文聚焦于基于「对话动作」（talk moves）框架的数学辅导话语分析，该框架源于负责任对话理论。然而，大规模辅导对话的收集、标注与分析在构建机器学习模型时面临资源密集型挑战。为此，我们提出 SAGA22 小型数据集，并探索多种建模策略，包括对话上下文、说话人信息、预训练数据集及微调方法。通过复用面向课堂教学设计的现有数据集与模型，实验表明在课堂数据上进行额外预训练可显著提升模型在辅导场景的表现，尤其当结合较长上下文与说话人信息时效果更优。此外，我们进行了广泛的消融实验，揭示了对话动作建模中的关键挑战。
原文摘要 · Abstract (English)
Human tutoring interventions play a crucial role in supporting student learning, improving academic performance, and promoting personal growth. This paper focuses on analyzing mathematics tutoring discourse using talk moves - a framework of dialogue acts grounded in Accountable Talk theory. However, scaling the collection, annotation, and analysis of extensive tutoring dialogues to develop machine learning models is a challenging and resource-intensive task. To address this, we present SAGA22, a compact dataset, and explore various modeling strategies, including dialogue context, speaker information, pretraining datasets, and further fine-tuning. By leveraging existing datasets and models designed for classroom teaching, our results demonstrate that supplementary pretraining on classroom data enhances model performance in tutoring settings, particularly when incorporating longer context and speaker information. Additionally, we conduct extensive ablation studies to underscore the challenges in talk move modeling.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。