Classification of Mobile Interactions Based on Human Emotion

Today, smartphones have become an integral part in one’s life. There are now techniques to identify the user’s emotional state as per their usage of the mobile device. In this paper, we present an attempt to classify the various interaction types such as scroll, tap, swipe, and pinch zoom based on human emotions. We run a formative statistical study where we have collected data from 60 participants over a period of 1 month. The participants were given different situations while using social networking applications and video entertainment applications and were asked to rate on 12 emotions (5 positive, 2 neutral, and 5 negative). Through the study, we find that there is significant relationship between different interaction types with satisfaction, surprise, and entertainment in emotions (p ≤ 0.05) for social networking applications. Also, there is significant relationship between different interaction types with satisfaction (p ≤ 0.05) in emotions for video entertainment applications. This result can be used while developing an interaction system. We could also develop a questionnaire that measures interaction types and emotions. The questionnaire reliability was tested using Cronbach’s Alpha.