Full-text resources of CEJSH and other databases are now available in the new Library of Science.
Visit https://bibliotekanauki.pl

Results found: 5

first rewind previous Page / 1 next fast forward last

Search results

Search:
in the keywords:  morfologické značkování
help Sort By:

help Limit search:
first rewind previous Page / 1 next fast forward last
1
Content available remote

Homonymie mezi oikonymy a antroponymy zakončenými na -slav/-slava jako problém automatické morfologické analýzy

100%
Osolsobě K., Žižková H.
Acta onomastica
|
2021
|
vol. 62
|
issue 2
377-403
EN
Homonymy at all levels, which is a distinct feature of all natural languages, is also one of the most significant obstacles to automatic natural language processing. In this paper, we will point out the morphosyntactic differences of Czech anthroponyms ending in -slav ( Miroslav-type, masculine) and Czech oikonyms with the same ending (Miroslav-type, feminine) and Czech anthroponyms ending in -slava (Miroslava-type feminine, because its forms are homonymous with both: masculine anthroponyms and feminine oikonyms). The analysis of data from the Syn v8 corpus shows that word form homonymy significantly influences the results of automatic morphological analysis. We will document errors in the coverage of the automatic analyzer dictionary and, above all, errors in morphological tagging, and we will propose a solution to partially improve the automatic disambiguation of the given type of proper nouns.
2
Publication available in full text mode
Content available

Korpusy jako zdroje dat pro úpravy nástrojů automatické morfologické analýzy (Slovotvorné varianty adjektiv na [(ou)|í]cí z hlediska morfologického značkování)

100%
Osolsobě K.
Časopis pro moderní filologii (Journal for Modern Philology)
|
2015
|
vol. 97
|
issue 2
136-145
EN
Adjectives ending with -oucí/-ící are regularly derived from verbs and hence are not usually listed in any of the Czech monolingual dictionaries. On the level of automatic morphological analysis (the dictionary) of Czech they should be generated from verbal roots and tagged as verbal adjectives (pos tag: AG.*). The data from Czech corpora prove a) inconsistencies in tagging and b) gaps in the dictionary. The main cause of both kinds of insufficiency is the existence of variants on the level of verbal forms from which the verbal adjectives are potentially derived. Consequently, text corpora are a significant source of knowledge about the formation and use of adjectives with endings -oucí/-ící that can be important for both a) automatic morphological analysis of Czech and b) theoretical description of Czech grammar (derivational morphology). Our goal is to present a corpus-based study of the Czech gerund, i.e. verbal adjectives with -oucí/-ící. The link between the inflected and the word-formation variants will be demonstrated using material from the SYN corpus (2,6 billion tokens of written Czech) and the large web corpus czTenTen12 (5,2 billion tokens of Czech text from the Internet — cleaned and deduplicated).
3
Publication available in full text mode
Content available

Korpus českého jazyka 2. poloviny 19. století

100%
Kučera K., Najbrtová K., Pivoňková K., Řehořková A., Stluka M.
Časopis pro moderní filologii (Journal for Modern Philology)
|
2019
|
vol. 101
|
issue 1
92-98
EN
The paper describes the principles and structure of the one-million-word DIA1900 Corpus built at the Institute of the Czech National Corpus (CNC) in Prague, focused on the language of Czech texts published in the years 1851 to 1900. The DIA1900, planned for publication by June 2020 and to be followed by the DIA1850 (a corpus built around the same principles, with the focus on the first half of the 19th century), observes both the balanced representation of the three major text types (belles lettres — journalistic texts — technical/scientific texts) and the system of morphological tagging implemented in the synchronic corpora included in the CNC project, thus facilitating the diachronic comparison of two stages in the development of Czech. A brief description is given of the structure of the morphological terminology used in the lemmatisation and tagging of the corpus, and of two tools designed to help search the 19th century texts with their fluctuating orthographic consistency combined with phonological and morphological variation characteristics of the language of the period: (1) a multiple select/suggest feature (reminding the user of the existence of non-standard orthographic and phonological variants of the lemma found in the corpus before the lemma search is started) and (2) the position attribute (informing the user of the ambiguous status of a word in the text, resulting from a misprint or misspelling, damaged page etc.).
4
Publication available in full text mode
Content available

Korpus DIA1900: jeho koncepce a vytváření

100%
Benešová L., Kučera K., Najbrtová K., Pivoňková K., Stluka M.
Časopis pro moderní filologii (Journal for Modern Philology)
|
2023
|
vol. 105
|
issue 1
121-140
EN
The objective of the paper is to describe the principles for building the onemillionword DIA1900 Corpus consisting of Czech texts published between 1851 and 1900, designed to be both balanced and representative. There are two main goals determining the methods of corpus building and the decision to develop new tools tailored to the special needs of 19th century Czech: 1) to present the variability of Czech in the 2nd half of the 19th century (including spelling, morphology, wordformation) and 2) to link the detected variants to the appropriate lemmas. The paper presents the phases of the processing of the texts, including transcription, manual pre-annotation, as well as the construction of a large morphological dictionary and the selection of a suitable set of paradigms. Further sections are focused on annotation and morphological tagging and manual disambiguation. The objective was to create a gold standard, intended for use in the automatic annotation both of the DIA1900 corpus and the planned corpus of Czech texts of the years 1800–1850.
5
Content available remote

Hluboké učení v automatické analýze českého textu

100%
Straková J., Straka M., Hajič J., Popel M.
Slovo a slovesnost: časopis pro otázky teorie a kultury jazyka (Slovo a slovesnost: A journal for the theory of language and language cultivation)
|
2019
|
vol. 80
|
issue 4
306-327
EN
The deep learning methods of artificial neural networks have seen a significant uptake in recent years, and have succeeded in overcoming and advancing the success of auto-solving tasks in many fields. The field of computational linguistics and its application offshoot, natural language processing, with classic tasks such as morphological tagging, dependency analysis, named entity recognition and machine translation, are no exception to this. This paper provides an overview of recent advances in these tasks related to the Czech language and presents completely new results in the areas of morphological marking and recognition of named entities in Czech, along with a detailed error analysis.
first rewind previous Page / 1 next fast forward last
JavaScript is turned off in your web browser. Turn it on to take full advantage of this site, then refresh the page.