Automatic classification of OGLE variables

Scientific exploitation of large variability databases can only be fully optimized if these archives contain, besides the actual observations, annotations about the variability...