Abstract
The objective of this study is to shed more light on the dependence between the performance of WSD feature-based classifiers and the specific features that may be chosen to represent a word context. In this paper we show that the features commonly used to discriminate among different senses of a word (words, keywords, POS tags) are overly sparse to enable the acquisition of truly predictive rules or probabilistic models. Experimental analysis demonstrates (with some surprising result) that the acquired rules are mostly tied to surface phenomena occurring in the learning set data, and do not generalize across hyponimys of a word nor across language domains. This experiment, as conceived, has no practical application in WSD, but clearly shows the positive influence of a more semantically oriented approach to WSD. Our conclusion is that feature-based WSD is at a dead-end, as also confirmed by the recent results of Senseval 2001.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Basili, R., Pazienza M.T., Velardi P.: An Empyrical Symbolic Approach to Natural Language Processing. Artificial Intelligence (1996)
Daelemans W., Van Den Bosch A. and Zavrel J.: Forgetting Exceptions is Harmful in Language Learning. Machine Learning 34 (1999) 11–43
Fukunaga K.: Introduction to Statistical Pattern Recognition. Academic Press, NY (1990)
Lee C. and Landgrebe D.A.: Feature Extraction Based on Decision Boundaries. IEEE PAMI 15,4 (1993)
Magnini B., Strapparava C., Pezzulo G. and Gliozzo A.: Using Domain Information for Word Sense Disambiguation. Proc. of SenSeval 2, Toulouse, France, (2001)
R. Mihalcea and D. Moldovan: An Automatic Method for Generating Sense Tagged Corpora. Proc. of AAAI-1999, Orlando, Florida, (1999)
R. Mihalcea and D. Moldovan: Pattern Learning and Active Feature Selection for Word Sense Disambiguation. Proc. of SenSeval 2, Toulouse, France, (2001)
Pederson T.: A Decision Tree of Bigrams is an Accurate Predictor of Word Sense. Proc. of NAACL, Pittsburgh (2001)
Velardi P. and Cucchiarelli A.: Theoretical Analysis of Context-based Learning Algorithms for Word Sense Disambiguation. Proc. of ECAI-2000, Berlin (2000)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cucchiarelli, A., Velardi, P. (2002). Feature-Based WSD: Why We Are at a Dead-End. In: Ranchhod, E., Mamede, N.J. (eds) Advances in Natural Language Processing. PorTAL 2002. Lecture Notes in Computer Science(), vol 2389. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-45433-0_3
Download citation
DOI: https://doi.org/10.1007/3-540-45433-0_3
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-43829-8
Online ISBN: 978-3-540-45433-5
eBook Packages: Springer Book Archive