Date
AI Overview
Why it matters
Adapting teaching slides to instructors' situated pedagogical needs is critical yet time-consuming, and current AI slide tools generate new presentations rather than refining existing slides to meet professional teaching standards.
Study at a glance
- Question: Can a multi-agent AI framework accurately and efficiently adapt existing teaching slides based on instructors' situated pedagogical needs?
- Setting and data: Five interviewed educators across primary, secondary, and tertiary levels; evaluation on 8 real-world slide decks with 16 modification requests.
- Method: Semi-structured interviews with thematic analysis, multi-agent prototype development, and evaluation against expert benchmarks and LLM scoring.
- AI tools: GLM Slide/Poster Agent, Office-PowerPoint-MCP-Server, Gemini-2.5-Pro
Key findings
- The multi-agent framework achieved an F1 score of 0.89 in operational agreement with expert-defined modification sequences, outperforming baseline tools.
- Automated evaluation showed the system scored highest in intent alignment (3.75 out of 5), content coherence (2.78), and factual accuracy (2.38) compared to baselines.
- The framework adapted slides in an average of 159 seconds, operating faster than the GLM agent but slower than the PPT-MCP baseline.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
The adaptation of teaching slides to instructors' situated teaching needs, including pedagogical styles and their students' context, is a critical yet time-consuming task for educators. Through a series of educator interviews, we first identify and systematically categorize the key friction points that impede this adaptation process. Grounded in these findings, we introduce a novel multi-agent framework designed to automate slide adaptation based on high-level instructor specifications. An evaluation involving 16 modification requests across 8 real-world courses validates our approach. The framework's output consistently achieved high scores in intent alignment, content coherence and factual accuracy, and performed on par with baseline methods regarding visual clarity, while also demonstrating appropriate timeliness and a high operational agreement with human experts, achieving an F1 score of 0.89. This work heralds a new paradigm where AI agents handle the logistical burdens of instructional design, liberating educators to focus on the creative and strategic aspects of teaching.
Research Categories
What is the application?
Who is the user?
Why use AI?
Study design
AI tools
Publication Source
Publisher
arXiv
