The main purpose of this research is using Brain Computer Interfaces (BCI) in the development of a model to characterize mental states using EEG signals, by implementing an automatic classification algorithm, and also by extracting features using the Fast Fourier Transform (FFT). Some experiments were carried out with 35 people. In the experiments, the relaxation and concentration states were classified using the SMO algorithm. The best result was an accuracy of 94.11%.

Keywords : EEG; BCI; OpenViBE; SMO; Mental States.