Back to Main Conference 2004
LREC 2004main
Semantic Annotating of Czech Corpus via WSD
Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC 2004)
Abstract
We would like to describe the relationship between word sense disambiguation (WSD) and language resources (LR) working with word senses. We discuss the problem of sense division and tagging. Exploiting specific features of the inflectional languages for WSD is encouraged. We present WSD methods for Czech ambiguous nouns. The advantage of these methods consists in reducing the manual work by using a synonym training set. They could be utilized for building a semantically annotated corpus or gaining glosses for Czech WordNet.