UD_Czech-CAC

Treebank card from UD release 2.18: metadata and texts come from UD_Czech-CAC/README.md; size is counted by streaming the .conllu files and checked against the release stats.xml. Generated by tools/treebanks build; do not edit by hand. Papers: Papers: UD_Czech-CAC; seeds: seeds/; converter: convert.md.

Field	Value
UD name	UD_Czech-CAC
Language	Czech (`cs`)
Family	IE
Treebank code	`cs_cac`
UD page	universaldependencies.org/treebanks/cs cac/index.html
License	CC BY-SA 4.0
Genres	news, nonfiction, legal, reviews, academic
Text in data	yes
Parallel	no
In UD since	UD v1.3
Authors	Barbora Hladká; Daniel Zeman
Contact	zeman@ufal.mff.cuni.cz
Development	outside the UD repo (elsewhere): fixes go to the source or the converter

Annotation layers

Layer	Origin
Lemmas	converted from manual
UPOS	converted from manual
XPOS	manual native
Features	converted from manual
Relations	converted from manual

UD scale: manual native — annotated by hand directly in UD; converted from manual — converted from manual annotation in another scheme; converted with corrections — converted and partly corrected by hand; automatic with corrections — automatic, with partial manual corrections; automatic — automatic; not available — the layer is absent.

Size

Split	Sentences	Tokens	Words	Multiword tokens	Empty nodes
train	23 478	471 594	472 609	1 015	799
dev	603	10 888	10 912	24	10
test	628	10 824	10 862	38	20
total	24 709	493 306	494 383	1 077	829

Tokens are surface tokens (Spanish vámonos is one), words are syntactic words (vamos + nos); multiword tokens are those split into several words; empty nodes are the N.M nodes of the enhanced graph.

Summary

The UD_Czech-CAC treebank is based on the Czech Academic Corpus 2.0 (CAC; Český akademický korpus; ČAK), created at Charles University in Prague.

Introduction

CAC consists both of written data and transcripts of spoken language. Only the written part is included in this treebank as no syntactic annotation is available for the spoken data. Out of 650,000 total CAC tokens, 493,306 appear in the treebank.

The first version of CAC was created by a team from the Institute of the Czech Language, Czechoslovak Academy of Sciences, led by Marie Těšitelová, in 1971-1985; its original name was “Korpus věcného stylu”. It was reshaped and made compatible with the Prague Dependency Treebank between 2007 (CAC 1.0) and 2008 (CAC 2.0); these corpora are distributed by the Linguistic Data Consortium. The corpus has now been converted to Universal Dependencies and made freely available under the CreativeCommons license (see LICENSE.txt).

See the following websites for more information on CAC 2.0:

ufal.mff.cuni.cz/rest/cac.html

ufal.mff.cuni.cz/rest/CAC/doc-cac20/cac-guide/eng/html/index.html

hdl.handle.net/11372/LRT-1061

catalog.ldc.upenn.edu/LDC2008T22

lindat.mff.cuni.cz//services/pmltq/#!/treebank/cac20/help (online query)

CAC contains mostly unabridged articles taken from a wide range of media. These articles include newspapers, magazines and other sources covering administration, journalism and scientific fields. These three genres can be distinguished by the sentence id: in
# sent_id = a-s20w-s55
the "s20w" part identifies the source document, where "w" means "written", "20" is the document id number and "s" means scientific (while "a20w" is the twentieth document from the administrative genre, and "n20w" from newspapers).

The texts are taken from the 1970s and 1980s and thus, the selection of texts is influenced by the political and cultural climate of this time period.

The original data in “Korpus věcného stylu” omits all punctuation symbols, numbers (only those expressed in digits; numeral words are preserved) and some measure units and symbols. The missing tokens were manually restored in CAC; however, for numbers and units, only a wildcard was inserted as the exact value and form cannot be guessed without access to the primary document.

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/treebanks/cs/cac/