Personalised Feedback Framework for Online Education Programmes Using Generative AI

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

2024-10-15

Abstract:AI tools, particularly large language modules, have recently proven their effectiveness within learning management systems and online education programmes. As feedback continues to play a crucial role in learning and assessment in schools, educators must carefully customise the use of AI tools in order to optimally support students in their learning journey. Efforts to improve educational feedback systems have seen numerous attempts reflected in the research studies but mostly have been focusing on qualitatively benchmarking AI feedback against human-generated feedback. This paper presents an exploration of an alternative feedback framework which extends the capabilities of ChatGPT by integrating embeddings, enabling a more nuanced understanding of educational materials and facilitating topic-targeted feedback for quiz-based assessments. As part of the study, we proposed and developed a proof of concept solution, achieving an efficacy rate of 90% and 100% for open-ended and multiple-choice questions, respectively. The results showed that our framework not only surpasses expectations but also rivals human narratives, highlighting the potential of AI in revolutionising educational feedback mechanisms.

Human-Computer Interaction,Artificial Intelligence,Cryptography and Security

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenge of providing personalized feedback in online education projects. Specifically, the author aims to improve the existing educational feedback systems by integrating generative artificial intelligence (Generative AI), especially by using large - language models such as ChatGPT. The goal of the paper is to develop a framework that can extend the capabilities of ChatGPT through embedding techniques, thereby achieving a deeper understanding of teaching materials and providing topic - specific feedback for quiz - based assessments.

### Core Problems of the Paper
1. **Improving the Quality and Timeliness of Feedback**: Traditional educational feedback systems usually have latency problems, which can impede the students' learning process. The method proposed in the paper aims to improve this situation by providing instant and personalized feedback through generative AI.
2. **Personalized Learning Experience**: By analyzing students' answer situations and learning behaviors, generative AI can provide more personalized feedback to help students better understand and master the course content.
3. **Automation and Teacher Support**: This framework can not only automatically generate feedback but also assist teachers in assessment and teaching, reducing the teachers' workload.

### Main Contributions
- **Framework Design**: The paper proposes a method that combines embedding techniques and Retrieval - Augmented Generation (RAG) to improve ChatGPT's performance in educational feedback.
- **Experimental Verification**: Through practical applications in the online cybersecurity education project, the effectiveness of this framework has been proven. The experimental results show that for open - ended questions and multiple - choice questions, the framework has achieved efficiency rates of 90% and 100% respectively.
- **Combination of Theory and Practice**: The paper not only provides theoretical discussions but also describes in detail the implementation methods and technical details of the system, providing references for other researchers and educational institutions.

### Conclusion
The paper demonstrates the great potential of generative AI in the education field, especially in providing personalized feedback. By integrating advanced natural language processing techniques, this framework can not only improve the quality and timeliness of feedback but also provide students with a more personalized learning experience, thereby enhancing the overall educational effect.