AI-based chatbot

• Main Author: Poon, Xing Wei
• Other Authors: Vidya Sudarshan
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/172097
• Institution: Nanyang Technological University

In recent years, the NTU website has witnessed significant growth, leading to challenges in information navigation, particularly for newly admitted students. This problem is exacerbated by the inherent ambiguity of natural language, making it exceedingly difficult to develop computer software that comprehends it. However, leveraging recent advances in deep learning and neural networks, AI chat bots have evolved from simple rule-based systems into sophisticated natural language processing (NLP) models capable of understanding and generating human-like text. To address these challenges, we present a chatbot model developed using these advanced NLP techniques, trained on a diverse set of potential user inquiries aligned with commonly asked questions. In this report, we delve into the current implementation of the question intent classification model within the chatbot, discussing both its advantages and disadvantages. This chatbot aims to enhance the user experience on the NTU website by providing efficient and contextually relevant information retrieval, ultimately easing the navigation and search process for all users, especially newcomers.