The influence of error types on the user experience of chatbots.
Keywords
Loading...
Authors
Issue Date
2020-07-08
Language
en
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Chatbots are being used and developed increasingly often in recent years. They can be
used in different areas, ranging from recommendation systems to customer support for
companies. Although the technology of such conversational agents is being used
increasingly often, such technology is not new. Dating back to 1950, when Alan Turing
developed a test for such agents. Another technology that is often used is that of
recommendation systems, which recommend items to users. Examples of areas where
these systems are used are areas like movie recommendation, for example in Net
ix. Or for music recommendation, like for example Spotify. This thesis focuses on different properties of a chatbot built for movie recommendation. Various mistake types were implemented and categories of the technology acceptance model (TAM) were evaluated for these chatbots. We found no hard conclusions due to lack of data. However, the data we collected shows a negative influence of the mistake types on the user experience categories that were evaluated. Aside from this, the relations between categories of the technology acceptance model were evaluated. We also added our own extension of the model, namely interactional enjoyment. We found that this category, just like ease of use and usefulness, correlates with the attitude towards using a chatbot. All these conclusions are not certain, since we lack scienti c power.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen