Abstract
In the modern society, damage of abuse of PINs (personal identification numbers) and passphrases will be recognized as social problem. One of these damages is a 'shoulder-surfing' attack on operating of bank ATMs to steal ATM users ' PINs. These snooping attacks to us must be opposed constantly for our safety. Not only bank ATM but also new mobile devices like a smart phone and tablet PC have to be consistently protected. In this article, we propose a rhythm authentication method using touch rhythms on the new devices such as smart phones and tablet PCs, and also we explain our analysis method with SOM (Self-Organizing Maps) as a suitable biometric authentication method. We experimented for three types of subjects - 1) the experienced person of the piano performance over 10 years, 2) students belong to department of information and computer sciences, and 3) the others - and we verified whether significant differences occurred. As the result, we obtained 99.2% precision in case of the experienced person of piano performance and then our method is simple, compact, and will be useful for smart devices also.
