Word Sense Disambiguation
Important Notions
- Word sense disambiguation based on Bayesian classifiers:
argmaxs P(s|C) = argmaxs P(s) P(C|s)
- Supervised learning
- Bilingual corpora
- Thesaurus classes
- Unsupervised learning
- Combining statistics with a priori constraints:
- One sense per collocation
- One sense per discourse
- Word sense discrimination (clustering)
Slides for lecture 8
Suggested Reading
- Krenn, B. & Samuelsson, C. (1997) The Linguist's
Guide to Statistics. Section 5.4-5.5.
- Manning, C. D. & Schütze, H. (1999) Foundations of Statistical
Natural Language Processing. MIT Press. Chapter 7.
- Yarowsky, D. (1995)
Unsupervised
Word Sense Disambiguation Rivaling Supervised Methods.
In Proceedings of the 33rd Annual Meeting of the Association
for Computational Linguistics. Cambridge, MA, pp. 189-196.
Project