Usability comparison of text CAPTCHAs based on english and Chinese

Abstract

Text CAPTCHAs are widely deployed in nowadays for websites to defend malicious attacks. Although most text CAPTCHAs employ alphanumeric characters, there are emerging interests in designing CAPTCHAs based on regional languages. Here, we conducted experiments to compare the usability of CAPTCHAs based on English and Chinese. The results indicate that, comparing with CAPTCHAs that employ random English or Chinese characters, those based on frequently-used English or Chinese words provide the best usability in terms of efficiency, effectiveness and satisfactory for participants who are native Chinese speakers while familiar with English. CAPTCHAs based on random Chinese characters, however, is least user-friendly from a comprehensive perspective. The evaluation method and results presented here may shine a light for the design of CAPTCHAs that employ characters other than alphanumeric.