Summary
The Slovak UD treebank is based on data originally annotated as part of the Slovak National Corpus, following the annotation style of the Prague Dependency Treebank.
Introduction
Slovak Dependency Treebank (Slovenský závislostný korpus) was created as part of the Slovak National Corpus at the Ľ. Štúr Institute of the Slovak Academy of Sciences. The original annotation follows the guidelines of the Prague Dependency Treebank (Czech), slightly modified in the spirit of Slovak grammatical tradition. Morphological tags, lemmas and dependency relations have been assigned manually to every word.
The present dataset is a subset of the original treebank. We automatically selected the sentences where the two human annotators 100% agreed on the analysis. This increases the quality and trustworthiness of the data but it also results in selecting short sentences most of the time. An extended version may be published in the future when manually merged and checked annotation is available.
This subset annotated in the original PDT-like style is available separately, see http://hdl.handle.net/11234/1-1822 and cite as
Gajdošová, Katarína; Šimková, Mária et al., 2016, Slovak Dependency Treebank, LINDAT/CLARIN digital library at the Institute of Formal and Applied Linguistics, Charles University in Prague, http://hdl.handle.net/11234/1-1822.
UD_Slovak contains the same data with annotation converted to conform to the Universal Dependencies guidelines. The original treebank was prepared by a team led by Katarína Gajdošová and Mária Šimková. Selection of sentences for this subset and conversion to Universal Dependencies was done by Dan Zeman.
References
- Daniel Zeman (2017): Slovak Dependency Treebank in Universal Dependencies. In: Jazykovedný časopis / Journal of Linguistics, ISSN 0021-5597, vol. 68, no. 2, pp. 385-395
Changelog:
- 2020-05-15 v2.6
- Changed sentence ids to preserve the name of the source text.
- Side effect: The order of the sentences changed, and so did the train-dev-test data split.
- Genitive, dative and instrumental nominals are now considered oblique.
- Added enhanced relations with case information.
- Added enhanced relations around relative clauses.
- Added enhanced external subjects in control verb constructions.
- Added empty nodes to enhanced graphs (but orphans are just converted to dep).
- Changed sentence ids to preserve the name of the source text.
- 2019-05-15 v2.4
- Fixed some bugs identified by the new validator.
- 2018-04-15 v2.2
- Repository renamed from UD_Slovak to UD_Slovak-SNK.
- Added enhanced representation of dependencies propagated across coordination. The distinction of shared and private dependents is derived deterministically from the original Prague annotation.
- 2017-11-15 v2.1
- Prepositional objects are now “obl:arg” instead of “obj”.
- Instrumental phrases for demoted agents in passives are now “obl:agent”.
- 2017-03-01 v2.0
- Converted to UD v2 guidelines.
- Improved advmod vs. obl distinction.
- Verb negation no longer treated as derivational morphology.
- Distinguished numeral types.
- Distinguished pronouns, determiners and pronominal adverbs.
- Distinguished coordinating and subordinating conjunctions.
- L-participles are verbs, other participles are adjectives.