MODELING UNIVERSITY ICT SERVICES AND SOLUTIONS USING AN ARTIFICIAL INTELLIGENCE CHATBOT

Abstract

Modeling university ICT services and solutions using AI Chatbot is building an AI-powered virtual agent to represent the human staff in providing similar services and solutions to ICT users online via Facebook Messenger and Web Demo platforms to enhance the accessibility of the ICT services. This study aims to model an artificial intelligence (AI) for university information and communication technology (ICT) services and solutions delivery by providing responses to questions asked by ICT service users. The modeled Chatbot was implemented to solve the challenges that traditional ICT staff in universities faced, such as high volume of requests, long wait times, numerous repeated questions, duty hours constraints, and inability to provide personalized support to ICT users at all times due to human natural capacities. To address these challenges, universities need to explore the use of AI Chatbot as a means of providing more efficient, personalized, and accessible ICT support services 24 hours per week (24/7). In this study, text and voice-based Chatbot was developed with Google Cloud Dialogflow and integrated it into Facebook Messenger and Web Demo for interactive text and voice messaging that enabled ICT service users to receive solutions as text or voice responses from Chatbot at anytime and anywhere. First, we identified 17 ICT services and classified them into 42 intents for building the Chatbot with 381 training phrases that formed the dataset. Second, 42 intents was created using Google Cloud Dialogflow Essential Console. Chatbot was trained for 3 months using rule-grammar and machine learning matching with 92% accuracy. The results of this study proved that the developed system can help universities provide more efficient and accessible ICT services to their ICT service users, as well as contribute to the advancement of AI Chatbot technology applications in the field of education.