University students often struggle to find reliable and unified sources of information for their queries, leading them to search for answers through informal channels such as friends, social media groups, or multiple university departments. This scattered approach results in misinformation, distractions, and an increased workload for university officials who receive repeated questions. Moreover, the Arabic language is frequently neglected in the field of Natural Language Processing (NLP), with limited research focused on developing effective models tailored for Arabic users. To address these pressing issues, we have developed an Arabic-language University Information Chatbot System. This chatbot consolidates a large repository of Frequently Asked Questions (FAQs), providing immediate, accurate, and accessible answers without the need for direct human intervention. Our research not only aims to enhance the efficiency of information dissemination across universities but also contributes to the advancement of Arabic language models in NLP. By utilizing state-of-the-art Arabic language models specifically designed for this purpose, our solution bridges the gap in Arabic NLP, which has historically received insufficient attention. The system ultimately benefits students, staff, and prospective students by reducing the time and effort needed to find information, fostering a more informed and self-sufficient academic community.

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

University Chatbot System Using Generative AI with Large Language Models

  • Aisha Ahmad,
  • Waad Alhejali,
  • Soumaya Chaffar,
  • Anis Koubaa

摘要

University students often struggle to find reliable and unified sources of information for their queries, leading them to search for answers through informal channels such as friends, social media groups, or multiple university departments. This scattered approach results in misinformation, distractions, and an increased workload for university officials who receive repeated questions. Moreover, the Arabic language is frequently neglected in the field of Natural Language Processing (NLP), with limited research focused on developing effective models tailored for Arabic users. To address these pressing issues, we have developed an Arabic-language University Information Chatbot System. This chatbot consolidates a large repository of Frequently Asked Questions (FAQs), providing immediate, accurate, and accessible answers without the need for direct human intervention. Our research not only aims to enhance the efficiency of information dissemination across universities but also contributes to the advancement of Arabic language models in NLP. By utilizing state-of-the-art Arabic language models specifically designed for this purpose, our solution bridges the gap in Arabic NLP, which has historically received insufficient attention. The system ultimately benefits students, staff, and prospective students by reducing the time and effort needed to find information, fostering a more informed and self-sufficient academic community.