Publication at Faculty of Mathematics and Physics |
2009
Abstract
As an important part of the Prague Dependency Treebank project a valency lexicon is being distributed. In this lexicon, more than 5300 verb entries are fully formally represented, with more than 8200 valency frames for verb senses included.
Moreover, the valency frames are interlinked with the Prague Dependency Treebank corpus, effectively providing a verb sense distinction and annotation for every occurrence of a verb in the corpus. More than 100,000 verb occurrences are annotated in this way.
The valency frame representation is fully formalized. In this contribution, we will concentrate on the formal description of the form of the verb