Automatic Prediction and Detection of Affect State Based on Invariant Human Computer Interaction and Human Physiological Response

In Human Computer Interaction (HCI) based on Web apps, the User Experience (UX) is of high importance for design specifications in industrial, commercial, academic purposes and UX investigations. Peering into the user's mind is of high value. We always like to understand user cognitive processes, why they do certain things online, and whether they see some elements on the webpage that we cannot. Methods such as interview and surveys normally gather this information by simply asking the users subjectively. Other methods such as the observational process mainly rely on viewing or observing the users' behaviour. In this paper we try to approach this in an objective perspective by directly observing what goes on in the user's mind, through non-obstructive technique without interrupting the data collection thereby obtaining qualitative and quantitative data of high dimension for modeling. Result shows high performance of models and a feasible solution for modeling HCI-HPR.