Human Resource Allocation in an Online Customer Service Chat System with AI-based Chatbot

碩士 === 國立成功大學 === 工業與資訊管理學系 === 107 === Customer service plays an important role between organizations and its customers. Among different types of customer service, a unique feature of customer service chat (CSC) is that agents can serve multiple customers simultaneously. The intelligent assistant (...

Full description

Bibliographic Details
Main Authors: Yi-ChengWang, 王逸程
Other Authors: Chung-Chi Hsieh
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/b7zgx8
Description
Summary:碩士 === 國立成功大學 === 工業與資訊管理學系 === 107 === Customer service plays an important role between organizations and its customers. Among different types of customer service, a unique feature of customer service chat (CSC) is that agents can serve multiple customers simultaneously. The intelligent assistant (Chatbot) that emerged along with the development of artificial intelligence has become an auxiliary tool for the CSC system. The organization imports the chatbot to the front interface of its CSC system for reducing the waiting time of customers and improving customer satisfaction. And the introduction of chatbot also affects the human resource allocation of the CSC system inherent in the organization. This study builds a CSC system that introduces a feedback-enabled chatbot through a simulation model and uses the simulation optimization tool to explore the human resource allocation strategy at different customer arrival rates. It also compares the total cost, as well as the percentage of abandonment before and after the import. The results of the simulation experiments revealed that in the case of low customer arrival rate, the introduction of the chatbot did not significantly reduce the average customer cost in the initial stage. In the long term, implementing chatbot brings significantly lower average total customer cost and the percentage of abandonment to the CSC system than the system without chatbots at three different arrival rates.