ProCAPTCHA: A profile-based CAPTCHA for personal password authentication

Abstract

CAPTCHA was introduced decades ago to distinguish between humans and bots. However, solving CAPTCHA has been a challenging issue for intruders. Various techniques, such as 3rd-party attacks, have been invented to break CAPTCHA. This research proposes ProCAPTCHA, a CAPTCHA system individually generated for each user by merging biometrics and user profiles. ProCAPTCHA leverages keystroke dynamics and personal information to create unique CAPTCHAs that are difficult for intruders to solve. ProCAPTCHA’s algorithm generates CAPTCHA based on the user’s profile data, ensuring randomness and uniqueness for each login. Performance evaluation shows that ProCAPTCHA can identify legitimate users with 100% accuracy, while only 60% of intruders are misclassified as true users. Bots face significant delays, often failing due to system time limits. Nonetheless, the bot’s attack must spend a very long time solving which, in real life, could be interrupted by the time limit of the system. Therefore, all bots cannot gain access as required.