Home // CENTRIC 2020, The Thirteenth International Conference on Advances in Human-oriented and Personalized Mechanisms, Technologies, and Services // View article
Authors:
Stephan Böhm
Judith Drebert
Sebastian Meurer
Olena Linnyk
Jens Kohl
Harald Locke
Levitan Novakovskij
Ingolf Teetz
Keywords: Chatbots; Conversational Design; Prototyping; User Intent Analysis; User-centered Design; Machine Learning
Abstract:
Chatbots are text-based dialogue systems that automate communication processes. Instead of communicating with a person, the user communicates with a computer system. Due to the use of Artificial Intelligence (AI) methods, such systems have become increasingly powerful in recent years and allow for more realistic dialogue processes. In particular, methods from the field of machine learning have contributed to an improved understanding of natural language. Nevertheless, such systems are not yet able to acquire the knowledge required to answer user queries independently. Dialogue structures and elements need to be defined as the conversational design of the chatbot. Herein, an user intent describes an information need or a goal that the user aims to achieve by entering text. For a user-centered chatbot design, a relevant set of intents must be identified and structured. In addition, training questions are required in order train the AI models for matching user input with the defined set of user intents. This article describes the procedure for developing chatbots using the example of an application in recruiting. The focus is on the appropriate identification and analysis of user intents. In our case study, the procedure for user-centered intent identification is described as well as approaches for the analysis and consolidation of intents. Furthermore, it is shown how corresponding measures affect the quality of intention identification.
Pages: 34 to 43
Copyright: Copyright (c) IARIA, 2020
Publication date: October 18, 2020
Published in: conference
ISSN: 2308-3492
ISBN: 978-1-61208-829-7
Location: Porto, Portugal
Dates: from October 18, 2020 to October 22, 2020