Towards Educator-Driven Tutor Authoring: Generative AI Approaches for Creating Intelligent Tutor Interfaces
Project Overview
This document explores the application of generative AI in education, specifically focusing on its use in designing user interfaces for Intelligent Tutoring Systems (ITSs). The research introduces a system leveraging Large Language Models (LLMs) and prompt engineering to automate the creation of tutor layouts and content. Built upon the Apprentice Tutor Builder (ATB), the system allows educators to specify requirements and generate interfaces at both the complete and component level. Preliminary evaluations indicate that this approach significantly improves efficiency in interface design, particularly for complex ITSs. The study highlights the potential of generative AI to streamline educational technology development by automating and accelerating the design process. The research also raises questions about integrating these tools into educator workflows and optimizing the balance between automated design and educator control.
Key Applications
AI-Enhanced Tutor Interface Builder
Context: Educators designing Intelligent Tutoring Systems (ITSs)
Implementation: Integrated into the Apprentice Tutor Builder (ATB), using an HTML/Javascript frontend, a Flask backend, and GPT-4 as the LLM. It uses a Domain Specific Language (DSL) and prompt engineering for interface generation.
Outcomes: Improved efficiency in tutor interface design. Preliminary evaluation showed a 23% reduction in time for a simple interface and a 68% reduction for a complex interface.
Challenges: Further research is needed to assess the quality and completeness of the AI-generated interfaces and their impact on the overall effectiveness of the tutoring system. Also, the optimal integration of generative interface tools into educators’ workflows, ensuring the system meets real-world tutoring requirements, is not fully addressed.
Implementation Barriers
Complexity of ITS development / Lack of educator design skills
The development of Intelligent Tutoring Systems (ITSs) often requires specialized programming and tutor design knowledge, hindering widespread adoption by educators who lack these skills. Existing authoring tools assume educators have the necessary skills to design effective and engaging tutor interfaces, which isn't always the case.
Proposed Solutions: The proposed AI-enhanced Tutor Interface Builder aims to address this by providing generative AI capabilities to assist educators in the UI design process.
Project Team
Tommaso Calo
Researcher
Christopher J. MacLellan
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Tommaso Calo, Christopher J. MacLellan
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gemini-2.0-flash-lite