Abstract
The pattern authentication method of Smartphone (e.g. Android lock pattern) has risks of being stolen
by the screen peeping (i.e. shoulder hacking) or being conjectured by residue on the screen. Accordingly, in
the previous experiment, we have tried to develop a security enhancement method for pattern authentication of
smartphone. To facilitate the user authentication in the same way as the pattern authentication, we applied
self-organizing maps (SOM) as a machine-earner/classifier for the authentication. It builds on both the picture-
password using locus and the angle of device holding. The result suggested the possibility to enhance the
security of pattern authentication. In this article, we verify practical use of an extended picture-password
method using behavioral profile and SOM for a biometrics authentication by the experiments.
