The Arab Open University, Oman is one of the premier institutions in the rapidly evolving educational landscape that embraces technology to assist students with support services. Artificial intelligence-driven chatbots are game changers as they can be used for customized guidance and support to change learning processes. However, recently, some studies have also focused on their ability to promote student engagement and self-directed learning and enhance academic performance. Approaching the problem traditionally sometimes leads to issues such as a lack of available advisors, inconsistency with information dissemination, and managing many students. These drawbacks complicate decision-making, course selection and academic performance. To meet this challenge, AOU started developing an AI chatbot-based student support academic advising system as the need for efficiency, responsiveness, and personalization of student support is growing. This project focuses on building an AI chatbot for an academic advisor. With state-of-the-art natural language processing (NLP), the bot will improve the advising experience for students and staff. For context-aware conversational interactions, we used the GPT API in the project and a simple Python script to make the local web interface for integrating the API. Because it is open-source, customizable, and scalable while also supporting individual user login, the customizability of OpenWebUI was highly attractive. We employed the AYA model from Cohere, which achieved 94% outstanding performance in serving queries on advising topics with the lowest resource utilization in Arabic and English.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Enhancing Academic Advising Through AI Chatbots: A Smart Support System for Students

  • Wasin Alkishri,
  • Mahmood Al-Bahri,
  • Yousuf Al Husaini

摘要

The Arab Open University, Oman is one of the premier institutions in the rapidly evolving educational landscape that embraces technology to assist students with support services. Artificial intelligence-driven chatbots are game changers as they can be used for customized guidance and support to change learning processes. However, recently, some studies have also focused on their ability to promote student engagement and self-directed learning and enhance academic performance. Approaching the problem traditionally sometimes leads to issues such as a lack of available advisors, inconsistency with information dissemination, and managing many students. These drawbacks complicate decision-making, course selection and academic performance. To meet this challenge, AOU started developing an AI chatbot-based student support academic advising system as the need for efficiency, responsiveness, and personalization of student support is growing. This project focuses on building an AI chatbot for an academic advisor. With state-of-the-art natural language processing (NLP), the bot will improve the advising experience for students and staff. For context-aware conversational interactions, we used the GPT API in the project and a simple Python script to make the local web interface for integrating the API. Because it is open-source, customizable, and scalable while also supporting individual user login, the customizability of OpenWebUI was highly attractive. We employed the AYA model from Cohere, which achieved 94% outstanding performance in serving queries on advising topics with the lowest resource utilization in Arabic and English.