Enhancing Library User Query Response Efficiency through AI Chatbot Development using Dialogflow Tools

Abstract

Khon Kaen University Library has implemented an automated question-and-answer system using the Dialogflow tool through the LINE Chatbot application. The KKU Library Chatbot has been used since November 2565 and has recorded statistics from November 2565 to August 2566, totaling 9 months. During this period, there were a total of 2,485 user interactions, averaging 276 interactions per month. Additionally, user satisfaction was evaluated to improve the LINE Official service to better meet user needs. The results from using the KKU Library Chatbot as a replacement for human operators in answering questions through the LINE Official channel showed a significant reduction in the number of human operators. Previously, there were 2 operators working 8 hours per day, but with the implementation of the 100% automated chatbot system, only 3 hours per month were needed for data verification and monitoring. Furthermore, comparing the number of answers provided by human operators via LINE with those provided by the chatbot in the same month of two different years (2022 and 2023), it was found that the chatbot provided three times more answers than the human operators. This demonstrates that the AI Chatbot can effectively replace human operators, assisting in answering repetitive questions continuously for 24 hours. These findings are consistent with the overall user satisfaction rating of 89.7% for the KKU Library Chatbot service.