An AI-based Solution for Enhancing Delivery of Digital Learning for Future Teachers
Project Overview
The document explores the innovative use of generative AI in education through the implementation of VidVersityQG, an AI-based solution that enhances digital learning and assessments for future teachers. By automating the creation of assessment questions from pre-recorded video lectures, it tackles the challenges of effectively gauging learner knowledge and competency. The design prioritizes a human-centered approach, enabling educators to modify the AI-generated questions, which fosters trust and engagement in the teaching process. The findings underscore the solution's efficacy in streamlining assessment workflows, leading to a significant reduction in the time educators traditionally spend on generating questions. Overall, the application of generative AI in this context not only improves the efficiency of educational assessments but also supports educators in delivering more personalized and effective learning experiences.
Key Applications
VidVersityQG
Context: Digital learning for future teachers in both university and corporate settings.
Implementation: The solution was implemented by automatically generating questions from transcribed educational video content using NLP techniques.
Outcomes: The solution showed high-quality question generation and significantly reduced the time and effort required for educators in manual question generation.
Challenges: Challenges included ensuring the accuracy of transcripts, generating subjective questions, and maintaining educator engagement.
Implementation Barriers
Technical
The accuracy of automatic speech recognition systems may lead to errors in transcription, affecting the quality of generated questions.
Proposed Solutions: Involve teachers in transcript inspection and correction to improve accuracy.
Pedagogical
The need for both objective and subjective question formats in assessments, with subjective questions being harder to generate automatically.
Proposed Solutions: Develop AI models that can generate both types of questions and allow teachers to modify them.
Trust and Engagement
Educators may be hesitant to trust AI-generated content without the ability to review and modify it.
Proposed Solutions: Implement a human-centered design that allows teachers to interact with and refine AI-generated questions.
Project Team
Yong-Bin Kang
Researcher
Abdur Rahim Mohammad Forkan
Researcher
Prem Prakash Jayaraman
Researcher
Natalie Wieland
Researcher
Elizabeth Kollias
Researcher
Hung Du
Researcher
Steven Thomson
Researcher
Yuan-Fang Li
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Yong-Bin Kang, Abdur Rahim Mohammad Forkan, Prem Prakash Jayaraman, Natalie Wieland, Elizabeth Kollias, Hung Du, Steven Thomson, Yuan-Fang Li
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai