Sometimes less is more : Romanian word sense disambiguation revisited

  • Recent approaches to Word Sense Disambiguation (WSD) generally fall into two classes: (1) information-intensive approaches and (2) information-poor approaches. Our hypothesis is that for memory-based learning (MBL), a reduced amount of data is more beneficial than the full range of features used in the past. Our experiments show that MBL combined with a restricted set of features and a feature selection method that minimizes the feature set leads to competitive results, outperforming all systems that participated in the SENSEVAL-3 competition on the Romanian data. Thus, with this specific method, a tightly controlled feature set improves the accuracy of the classifier, reaching 74.0% in the fine-grained and 78.7% in the coarse-grained evaluation.

Download full text files

Export metadata

Author:Georgiana Dinu, Sandra KüblerORCiDGND
Editor:Galia Angelova, Kalina Bontcheva, Ruslan Mitkov, Nicolas Nicolov, Nikolai Nikolov
Document Type:Preprint
Year of Completion:2007
Year of first Publication:2007
Publishing Institution:Universitätsbibliothek Johann Christian Senckenberg
Release Date:2008/11/03
Romanian; Word Sense Disambiguation; memory-based learning
Page Number:5
Erschienen in: Galia Angelova ; Kalina Bontcheva ; Ruslan Mitkov ; Nicolas Nicolov ; Nikolai Nikolov (Hrsg.): International Conference Recent Advances in Natural Language Processing : proceedings, Shoumen : Incoma, 2007, S. 173-177, ISBN: 978-954-91743-7-3
Source: ; Proceedings of the International Conference on Recent Advances in Natural Language Processing, RANLP 2007 - Borovets, Bulgaria, September 2007.
Institutes:keine Angabe Fachbereich / Extern
Dewey Decimal Classification:4 Sprache / 40 Sprache / 400 Sprache
Linguistik-Klassifikation:Linguistik-Klassifikation: Computerlinguistik / Computational linguistics
Licence (German):License LogoDeutsches Urheberrecht