The main scope of this dissertation is to identify and thus authenticate users based on
their behavior while playing a game in an android smart phone device. For this reason,
an android game, called RoboCombat, was utilized and edited to store the individual's
interacting movements. Subsequently, ten volunteers participated in the recording
phase and provided their data which contributed in forming their corresponding
profiles. The classification process uses these profiles and tries to predict the identity
of a player. Several experiments were made which ascertain the system's performance
on correctly authenticating a user.
In the beginning, the introduction section explains some basic terms in order for the
reader to be familiarized with the dissertation's concept. In more details, the reader
can obtain information regarding the biometric authentication process. Then, the
related work section presents some similar and interesting approaches on behavioral
authentication of a user. Next, the methodology part describes analytically the steps of
developing the authentication system along with the purpose of each step. In this way,
the experiments can be replicated. In the following chapter, the results of the
experiments are presented and visualized. Finally, the conclusion chapter summarizes
the system's performance.
Collections
Show Collections