Towards prediction of user experience from touch interactions with mobile applications

Carola Trahms, Sebastian Möller, Jan-Niklas Voigt-Antons

In: 19th International Conference, HCI International, Part II. International Conference on Human-Computer Interaction Cham, Switzerland Seiten 505-512 ISBN 978-3-319-58753-0 Springer International Publishing AG 2017.


In this work a first attempt is made to model user experience ratings with recordings of touch interactions. To measure the user experience the AttrakDiff Mini questionnaire is used providing three quality dimensions: pragmatic quality, hedonic quality and attractiveness. The feature selection for linear models shows that it is possible to predict the pragmatic quality for a test set with an r2 of up to 0.64. The hedonic quality and attractiveness, however, seem not to be reflected in these touch interactions in a linear way. Still, the results from this work constitute a promising direction for future research on assessing or even optimizing the user experience for apps automatically in real time.

Weitere Links

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence