Chatbot in the Service Industry: Challenges and Perspectives

Abstract

The chatbot is a software application that could initiate and participate in a meaningful conversation with a live human being agent. With proper algorithms and sufficient training, a chatbot could comprehend inquiries and replies in a way that mimics human response. The chatbot allows companies to operate at a low cost and provides more efficient and professional customer service. This article aims to provide a symmetrical overview of chatbot building and comments on frequently used models and mainly focuses on the models that researchers utilize to build chatbots. The models are separately discussed in terms of query understanding and response generation, and the knowledge base, evaluation, and optimization methods are also included. Chatbots building is a combination of Natural Language processes (NLP) and Machine Learning (ML), and there is a trend shifting from rule-based to more complex models involving Neural Networks.