Stress detection of computer user in office like working environment using neural network

Detecting the stress of computer user in an office like environment will enable more development of computer and make it intelligent enough where it can interact with its user, taking users effective state in to account; known as affective computing. In this research work, we have analyzed physical and mental stress of a computer user in all day long working environment by analyzing variations in physiological signals. Physiological data sets of 12 subjects were collected where all the subjects were went through a specific sequence of computer using session which includes different computer mediated task. To determine the stress level accurately and for detailed analysis of stress condition a three layer back propagation (BP) neural network were constructed. Research shows that stress level of a subject varies with computer using time and subject´s effort towards work. And induced stress is mainly due to intense eye work and mental strain.