A case study of the robustness and the usability of CAPTCHA

The websites and network application experienced explosive growth in the past two decades. As the evolution of smartphones and mobile communication network have evolved, smart phone s user experience has been improved to a high level, and more and more people prefer to use smartphones. However, the development of techniques will not only increase the users experience but also bring threats of cracking. The development of techniques brought the potential threats to websites security. As a result, CAPTCHA, Completely Automated Public Turing test to tell Computers and Humans Apart, forms one of the methods to impede spamming attacks. As CAPTCHA s definition indicates, CAPTCHA should be recognized by humans easily while shouldn t be recognized computers. These two attributes of CAPTCHA can be considered as usability and robustness. Some CAPTCHA is difficult to be recognized by computers, but humans may also find difficult to recognize it. Therefore, the purpose of the thesis is to find out the balance between usability and robustness of CAPTCHA. Therefore, the related researches about the usability and the robustness of CAPTCHA will be reviewed, and the process of automatic CAPTCHA recognition will be Figured out and implemented by the author. The implementation will be based on the existed algorithms and a case study. The findings are the factors for improving CAPTCHA s robustness. They are from the each step of a specific process of automatic CAPTCHA recognition. Then the factors will be compared with the issues which are from the related usability research. The discussion will derive some possible ways, such as adding confusing characters and increasing data s diversity to improve robustness while keeping the usability according to the derived factors