Skip to main content Skip to navigation

Personalised Feedback Framework for Online Education Programmes Using Generative AI

Project Overview

The document explores the integration of generative AI, specifically ChatGPT, in online education to improve personalized feedback within learning management systems (LMS). It outlines the creation of a feedback framework that leverages AI to deliver timely and individualized feedback to students, particularly in online courses such as cybersecurity. The findings indicate that AI-generated feedback is highly effective and can compete with human feedback, tackling common challenges of delayed and subpar feedback typically found in traditional educational environments. By utilizing learning analytics, the framework is designed to enhance student engagement and academic performance through tailored support, ultimately demonstrating the potential of generative AI to transform educational experiences and outcomes.

Key Applications

Automated Personalized Feedback Generation

Context: Various educational environments including online education programs, chemical engineering coursework, higher education writing assignments, and Moodle learning environments, aimed at enhancing student engagement and learning outcomes.

Implementation: Utilized AI technologies, including ChatGPT and Microsoft Power Tools, integrated with platforms like Moodle and Excel to automate the analysis of student submissions and provide personalized feedback based on performance data and specific student responses.

Outcomes: Achieved significant improvements in student performance, with efficacy rates of up to 90% for open-ended questions and an average increase of 23.3% in student marks. Enhanced engagement through tailored feedback.

Challenges: Faced initial discrepancies in marking accuracy, limitations in handling complex responses, dependence on existing IT infrastructure, and time constraints associated with traditional feedback methods.

Implementation Barriers

Technical Barrier

Integration complexity between AI models and existing LMS frameworks.

Proposed Solutions: Utilizing APIs and modular design for easier integration.

Quality Assurance Barrier

Ensuring the accuracy and relevance of AI-generated feedback.

Proposed Solutions: Implementing reference sets and continuous evaluation to refine feedback quality.

Ethical Barrier

Concerns about bias in AI feedback and its impact on student learning.

Proposed Solutions: Establishing guidelines for ethical AI use and ongoing monitoring for bias.

User Acceptance Barrier

Resistance from educators and students to adopt AI-based feedback tools.

Proposed Solutions: Providing training and demonstrating the effectiveness of AI tools in improving learning outcomes.

Project Team

Ievgeniia Kuzminykh

Researcher

Tareita Nawaz

Researcher

Shihao Shenzhang

Researcher

Bogdan Ghita

Researcher

Jeffery Raphael

Researcher

Hannan Xiao

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Ievgeniia Kuzminykh, Tareita Nawaz, Shihao Shenzhang, Bogdan Ghita, Jeffery Raphael, Hannan Xiao

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18