Automated detection of user deception in on-line questionnaires with focus on eye tracking use

On-line questionnaires are today widely used for various tasks, from census data collection to knowledge testing in job interviews. However, there is currently no automated system that can help us decide if the answers from the questionnaires are reliable or estimate how reliable the are. Deception is a part of everyday human behavior and deception is also present when answering on-line questionnaires. People are trying to make themselves look better or are just withholding information for malicious reasons. In our paper we present a method for automatic prediction of honesty for answers in a questionnaire. We demonstrate that by using new technologies like eye-tracking, we can create an automated system which can help us estimate reliability and truthfulness of the answers from on-line questionnaires. In our paper we have proposed and evaluated several metrics that can be used for automated detection of user deception in on-line questionnaires and we have also created and tested our first automated system for deception detection, based on these metrics.