Human-Cognition-Based CAPTCHAs
Human-Cognition-Based CAPTCHAs
复制标题
DOI：
10.1109/mitp.2015.79
复制
发表时间：
2015-09
期刊：
影响因子：
2.6
通讯作者：
S. Saha;A. Nag;D. Dasgupta
中科院分区：
文献类型：
--
作者：
S. Saha;A. Nag;D. Dasgupta
在任何基于web的身份验证过程中，captcha在区分人和机器方面发挥着重要作用。随着文本识别和图像提取技术的进步，现在可以以令人满意的准确性提取CAPTCHA中显示的字符。引入captcha是为了应对在线认证系统面临的巨大威胁，但对于普通互联网用户来说，解读它们变得越来越困难。作者的替代CAPTCHA提供了各种只有人类才能理解和正确回答的数学、逻辑和推理问题。该框架支持问题多样性和用户友好界面。作者通过非正式用户研究对不同背景下开发系统的绩效进行了评价。研究表明，所实施的系统效果良好，用户满意度较高。
CAPTCHAs play a significant role in differentiating humans and machines in any Web-based authentication process. With technological advances in text recognition and image extraction, it's now possible to extract the characters shown in a CAPTCHA with satisfactory accuracy. CAPTCHAs were introduced to cope with the immense threat to online authentication systems, but interpreting them becomes harder for regular Internet users every day. The authors' alternative CAPTCHA provides a variety of mathematical, logical, and inference problems that only humans can understand and answer correctly. The proposed framework supports question diversity and a user-friendly interface. The authors' informal user study evaluates the developed system's performance with different backgrounds. The study shows the efficacy of the implemented system with a good level of user satisfaction.