Application of multinomial mixture model to text classification

0411075 - UTIA-B 20030062 RIV DE eng C - Konferenční příspěvek (zahraniční konf.)
Novovičová, Jana - Malík, Antonín
Application of multinomial mixture model to text classification.
Berlin: Springer, 2003. Lecture Notes in Computer Science., 2652. ISBN 3-540-40217-9. In: Pattern Recognition and Image Analysis. - (Perales, F.; Campilho, A.), s. 646-653
[Iberian Conference on Pattern Recognition and Image Analysis. IbPRIA 2003 /1./. Puerto de Andratx (ES), 04.06.2003-06.06.2003]
Grant CEP: GA AV ČR IAA2075302; GA AV ČR KSK1019101
Výzkumný záměr: CEZ:AV0Z1075907
Klíčová slova: text classification * multinomial mixture model * Bhattacharyya distance
Kód oboru RIV: BB - Aplikovaná statistika, operační výzkum

The mixture of multinomial distributions is proposed as a model for class-conditional distributions in document classification task. Experimental results on the Reuters and the Newsgroups data sets indicate the effectiveness of the multinomial mixture model. Furthermore, an increase in classification accuracy is achieved for small training data sets, when multiclass Bhattacharyya distance is used instead of average mutual information as a feature selection criterion.
Trvalý link: http://hdl.handle.net/11104/0131162