Checking Internet masquerading using multiple CAPTCHA challenge-response systems

摘要

CAPTCHAs are challenges or tests administered automatically over networks that can distinguish between people and machines (automated script) and thus protect web services from abuse by programs disguising as human users. Most existing CAPTCHA challenges are a single challenge response system (CR) that provide ready-made answers to the users in which case a bot may answer the test and escape scot-free. In single response systems, a user is provided the options of either providing already made answer to the test or solving a simple mathematics question such ¿1+1¿ which requires the user to do some computation. In this paper, a Multiple Challenge-Response system call M-CR CAPTCHA mechanism is proposed. Here a user is expected to respond to multiple questions before access is granted; thus making the system more robust than existing single challenge-response systems. The system can thus be used to double checked against masquerading. Humans can solve the multiple challenge quickly and accurately while it will be computationally costly for a bot to solve it. Experiments on the developed system revealed that users who understand the purpose of CAPTCHA code advocate a dual layer CAPTCHA security mechanism and enjoy more robust security using the multiple CAPTCHA system.