Touch is an important means of expressing emotion in everyday life, but how do we express emotions in games played using a touchscreen? In this monograph, Gao et al. carefully investigate this question. The popularity of gaming in a touch-based environment, such as a smartphone or tablet, means that the results of this investigation may affect game design in general, and may result in reactive personalization for the player in particular. The authors use machine learning techniques to capture finger-stroke factors and assess four emotional states (excited, relaxed, frustrated, and bored), two arousal levels (low and high), and two valence levels (positive and negative).

After providing introductory motivation, the authors present a careful review of the literature on touch behavior and its connection to emotion. This is followed by the experimental protocol for collecting and labeling data to build and test their system to recognize emotion. A careful analysis of touch features and how they relate to player emotion leads into a discussion of the building and testing of models for automated recognition. The underlying classification algorithms include “discriminant analysis (DA), artificial neural network (ANN) with back propagation, and support vector machine (SVM) classifiers.” The paper concludes with a discussion of the results and how these techniques might apply in other contexts. Nine figures and nine tables provide details on the many technical aspects of this study, and 88 references indicate the careful preparation that went into this monograph.

Readers interested in human-computer interfaces, games, and app design will benefit from this well-organized, insightful, and technically precise publication.