An empirical pilot study of CAPTCHA complexity using eye tracking

Abstract

A "Completely Automated Public Turing test to tell Computers and Humans Apart" (CAPTCHA) is a widely used technique in many websites for protecting their online services from malicious users. Yet, end-user perceptions related to the complexity of CAPTCHA varies based on the visual properties of a CAPTCHA. In this paper we present an empirical pilot study that employs eye tracking to measure the complexity of CAPTCHA with different properties using three metrics: fixation count, fixation duration and task completion. Our initial results reveal that with high CAPTCHA complexity, the fixation count and fixation duration increase, on the other hand, the task completion decrease. The approach we used can be incorporated with other metrics for providing an additional technique to measure the usability of CAPTCHAs.