A novel unsupervised classifier of the P300 presence based on a match filter is proposed. With the combination of different artifact cancellation methods and P300 extraction techniques. This innovation brings a notable impact in ERP-based communicators. Database from a Donchin ERP-based speller is investigated.