Date:

2026-07-01

DOI Number:

10.1145/3766557

Institution:

人工智能教育学部

Title of Paper:

How does Agents reconfigure human-computer Interaction dialogues?– An Empirical Study of the Mechanism of Problem Scaffolding Action Based on Epistemic network Analysis

Journal:

Proceedings of the 2025 International Conference on Educational Technology and Artificial Intelligence

Key Words:

Smart Learning Companion, interactive human-computer dialog, Problem scaffolding, Epistemic network analysis

Summary:

In the context of the digital transformation of education, the development of generative AI is driving the reshaping of the teachinglearning interactive dialogue model. Although some studies have shown that it can improve the quantity and quality of student interactive dialogues, the characteristics of student-human-computer interactive dialog models are not yet clear. This study examines the effects of Problem scaffolding-supported Smart Learning Companion (SLC) on the educational dialog model and students’ TPACK competency development from an instructional design task, combined with Epistemic network analysis techniques. It was found that Problem scaffolding effectively stimulated students’ critical thinking and significantly increased questioning, reasoning, and perspective construction behaviors in conversations compared to no scaffolding support; while conversations without scaffolding mainly stayed at the shallow level of information exchange. This study provides methodological innovations and practical insights into the design of dialogue scaffolding for educational AI, which can be used in the future to develop an intelligent scaffolding adaptive system by combining multimodal data and adaptive technologies to further optimize deep learning paths in human-computer collaboration.

Co-author:

张屹

First Author:

周平红

Document Code:

1998638684886405120

Discipline:

教育学

Document Type:

00013

Translation or Not:

No

Date of Publication:

2025-11-27