Klasifikacija ključnih besed iz bibliografskih enot
A research field dealt with in a bibliographic source is not easily identifiable from the COBISS bibliographic data on Slovenian researchers. Keywords are the best way to identify research fields. Our aim was to sort (classify) keywords of researchers’ bibliographic units. To this end, we have chosen as the classifier the classification of the Slovenian Research Agency. A special attention has been devoted to the selection and preparation of texts intended for machine learning in order to achieve best results. To classify keywords from the researchers’ bibliographical units, we have used Oracle Data Miner and the support vector machines method which has already proven to be useful for the classification of texts.