Predicting personality traits from touchscreen based interactions

Abstract

This study explores whether unobtrusive features from everyday touchscreen interactions can predict stable personality traits. Using gesture dynamics (e.g., tap timing, swipe kinematics) gathered from mobile tasks, we trained models that achieved meaningful correlations with Big Five dimensions. Results highlight both the promise and ethical considerations of behavior-based personality inference in mobile UX research.

BibTeX

@inproceedings{voigtantons2018c45,
  author    = {Küster, L. and Trahms, C. and Voigt-Antons, J.-N.},
  title     = {Predicting personality traits from touchscreen based interactions},
  year      = {2018},
  booktitle = {Paper presented at the 10th International Conference on Quality of Multimedia Experience (QoMEX 2018), Sardinia, Italy},
  doi       = {10.1109/QoMEX.2018.8463375},
}