Version 1
: Received: 2 October 2024 / Approved: 3 October 2024 / Online: 4 October 2024 (08:10:44 CEST)
How to cite:
Al-Jaf, K.; Öz, C.; Rashid, T. A. Beyond ChatGPT: A Hybrid Chatbot Model for Reliable Educational Administration. Preprints2024, 2024100276. https://doi.org/10.20944/preprints202410.0276.v1
Al-Jaf, K.; Öz, C.; Rashid, T. A. Beyond ChatGPT: A Hybrid Chatbot Model for Reliable Educational Administration. Preprints 2024, 2024100276. https://doi.org/10.20944/preprints202410.0276.v1
Al-Jaf, K.; Öz, C.; Rashid, T. A. Beyond ChatGPT: A Hybrid Chatbot Model for Reliable Educational Administration. Preprints2024, 2024100276. https://doi.org/10.20944/preprints202410.0276.v1
APA Style
Al-Jaf, K., Öz, C., & Rashid, T. A. (2024). Beyond ChatGPT: A Hybrid Chatbot Model for Reliable Educational Administration. Preprints. https://doi.org/10.20944/preprints202410.0276.v1
Chicago/Turabian Style
Al-Jaf, K., Cemil Öz and Tarik A. Rashid. 2024 "Beyond ChatGPT: A Hybrid Chatbot Model for Reliable Educational Administration" Preprints. https://doi.org/10.20944/preprints202410.0276.v1
Abstract
The ever-growing student population and limited administrative resources strain traditional communication channels within educational institutions. Chatbot technology offers a promising solution, but current limitations can impede effective student-administration interaction. This paper proposes a hybrid chatbot model to enhance reliability in educational administration by addressing the limitations of generative models like ChatGPT. We investigate three individual chatbot approaches (rule-based, retrieval-based, and generative-based) to handle educational inquiries. An evaluation assessed their strengths, weaknesses, and user experience. Based on these findings, a hybrid model with an intelligent classifier for query routing was developed. This model leverages each approach's strengths and addresses uncertainty in the generative model to achieve increased accuracy and reliability. Our results demonstrate an accuracy improvement compared to individual models. This work contributes to developing adaptable chatbot systems capable of addressing a wider range of user inquiries, which is particularly beneficial in resource-constrained educational settings. Additionally, the proposed model demonstrates competitive performance when benchmarked against existing state-of-the-art administrative educational chatbots. Furthermore, by automating routine inquiries and offering 24/7 access to information and support, this model has the potential to reduce administrative workload and improve the overall student experience significantly. Moreover, the modular design of our hybrid model offers a foundation for future research on personalizing chatbot interactions and integrating domain-specific knowledge for enhanced communication within educational environments.
Keywords
Chatbots; ChatGPT; Rule-based; Retrieval-based; Generative-based; Educational Administration; Hybrid model
Subject
Computer Science and Mathematics, Computer Science
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
