A Cloud-Based Mobile Anger Prediction Model

Anger is a main human emotion mostly healthy for our life, but when it gets out of control, it can lead to problems. With the popularity of social media and smartphone devices, the problems shift into the network worlds. We have seen that many people cannot control their temper and result in destructive behavior happened on the interaction with cybercitizens. Not only harm their on-line friends but also destruct themselves on the social network. Hence, we hope to help smartphone users control their emotion in the network world and improve their emotional intelligence. To archive this goal, a cloud-based mobile anger prediction model has been proposed, and based on the specification of model a prototype has been constructed. The prototype includes a smartphone application agent and the cloud-side analytical service. The application agent use sensor fusion to transform emotion cues as data tuples. The cloud analytical service analyzes the submitted data tuples by the classifier and then utilize recommender model to predict the intensity of anger happening. After analyzing, the prediction messages have been pushed back to the agent. The agent uses this perdition to alert the mobile users for their emotion management during their social media interaction. The experimental result shows that the recommender model is well-design for the anger prediction. Moreover, with the advantages of scalability and mobility, the future extension on the mobile emotion management field is expectable.