Cognıtıve Stress Detectıon Usıng Keystroke Dynamıcs And Pattern Varıatıons

Abstract

Stress is caused due to certain performance parameters, daily routines, habitual situations, and other biological, internal, external factors in our being. Stress is a widespread phenomenon that is integrated knowingly or unknowingly in our day-to-day lives. This element of stress can prove highly disastrous, hazardous as it is potentially harmful and even deathly in certain cases. This research focuses on analysis of stress, detection of cognitive stress using the novel concept of keystroke dynamics and pattern variations. The keystrokes are collected using an Android application and stress factor is also induced using the same. There are around 13 visibly important and unique features that are collected per user. The model building is also performed using classification algorithms to predict the final output, that is, how stressed out an individual is at any given point of time.