AplikaceAplikace
Nastavení

Toto je starší verze dokumentu!


Korpus textů Václava Klause

Corpus Klaus ('VK') is an author corpus of Václav Klaus’ texts, which was created as a data basis for the thesis Václav Klaus’ Idiolect: A Corpus-based Analysis. The data used for the creation of the corpus were sourced from Klaus’ official website, which contains texts intended primarily for this website, as well as texts originally published elsewhere (e.g. newspaper articles or magazine interviews) or created for specific events (e.g., presidential speeches or lectures at conferences).

In addition to Klaus’ texts, the website also contains texts for which Klaus is only a co-author (e.g. joint statements) or for which he is not an author (e.g. communications from the press department of the presidential office). However, the 'VK' corpus is an author corpus in the narrower sense and, therefore, does not include these texts. For many texts, especially for a considerable portion of interviews, the mode (written or spoken) cannot be reliably determined. In the case of the spoken texts (the debates and some interviews), the situation is complicated by the apparent editorial modifications of Klaus’ speeches, the extent and nature of which vary considerably from text to text. To preserve the authenticity of the linguistic material, the corpus does not contain texts whose mode could not be clearly identified, nor does it include ‘purely’ spoken texts. The following four conditions can define the texts selected for the corpus:

  1. only texts published on the website www.klaus.cz;
  2. only texts whose sole (listed) author is Václav Klaus;
  3. only texts in the written and written-to-be-spoken (i.e. texts originally written but intended to be spoken) mode;
  4. only texts published up to and including 31st October 2023.

The entire corpus consists of 2,313 documents, which have been assigned the following 14 structural attributes:

structural attribute description (values)
doc.id document identification name (various values)
doc.title original title (various values)
doc.lang language (Czech)
doc.src_lang source language (Czech)
doc.author author (Václav Klaus)
doc.pubDateYear year of publishing (values in scope of 1995–2023)
doc.date date of publishing (various values)
doc.period office held by Klaus at the time of publishing (Prime Minister, MP, Speaker of the Chamber of Deputies, President, ex-president)
doc.modus mode (written, written-to-be-spoken)
doc.registr register (documents, professional literature, journalism, public speeches)
doc.txtype text type or genre (various values)
doc.medium type of the text source (various values)
doc.source text source (various values)
doc.comment commentary or additional information (various values)

The corpus size is 1,750,891 tokens in total and 1,475,640 tokens excluding punctuation. All tokens were assigned the following 11 positional attributes during automatic annotation:

positional attribute type count incl. punctuation type count excl. punctuation
word 98 892 98 830
lc 88 814 88 752
sforma 98 896 98 831
lemma 36 604 36 555
lemma_lc 35 893 35 844
sublemma 37 942 37 892
sublemma_lc 37 225 37 175
tag 2013 2007
pos 15 14
case 8 8
verbtag 77 77

How to cite

Schmid, O.: Idiolekt Václava Klause: korpusová analýza. Diplomová práce. Ústav českého jazyka a teorie komunikace FF UK, Praha 2024. Dostupný z WWW: <http://hdl.handle.net/20.500.11956/191695>