Slovak Morphosyntactic Tagset

Abstract

Morphological annotation constitutes essential, very useful and very common linguistic information presented in corpora, especially for highly inflectional languages. The morphological tagset used in the Slovak National Corpus has been designed with several goals in mind – the tags are compact and easily human-readable, without sacrificing their informational contents. The tags consist of ASCII letters, numbers and several other characters. In general, they have a variable number of symbols, but their order is obligatory, and each category or specific feature is assigned a particular character, which can be shared among several parts of speech. The tagset is highly functional and pragmatic, although some allowances had to be made to accommodate traditional analysis of Slovak morphology and part of speech categories. In particular, function words are classified according to their syntactic (and semantic) roles, which is a reason why the tagset is sometimes described as a morphosyntactic one.

Improvement of Misleading and Fake News Classification for Flective Languages by Morphological Group Analysis

“…The most important contribution to the automated morphological analysis of the Slovak language was a proposal of the morphologically annotated corpus of the by Slovak National Corpus. The corpus defines a set of Part-of-speech (POS) tags and annotation guidelines for the Slovak language, and it is still the most used set of morphological tags [25]. The tags are of unequal length, but most tags follow the same structure for the same inflectional paradigm.…”

Corpus of Slovak Legislative Documents

“…State-of-the-art lemmatization and tagging for Slovak is performed by the MorphoDiTa tagger (Straková et al 2014) that we trained on manually lemmatized and MSD annotated Slovak corpus r-mak-6.0 of 1.2 million tokens. 6 The accuracy of lemmatization on general texts is 98.2%; the accuracy of MSD tagging 94.2%; the accuracy of POS tagging 98.1% and the combined accuracy of lemmatization+MSD tagging is 93.5% (Garabík -Mitana 2022). Unfortunately, we cannot easily estimate the accuracy on the texts from the legal language domain; nevertheless, since the training corpus r-mak-6.0 contains one manually lemmatized and MSD annotated legal text, although strictly speaking not a law (Programové vyhlásenie vlády 7 ), we can obtain at least a rough estimate by training a separate tagger model on the rest of the corpus and calculating the accuracy on the one legal text.…”

Extraction and Presentation of Bilingual Correspondences from Slovak-Bulgarian Parallel Corpus

“…At the first step of our study we prepare morphologically annotated sentence-aligned parallel texts. The Slovak texts are morphologically annotated automatically by the tagger Morče which has been trained and tuned on tagset, developed by the Slovak National Corpus (Garabík & Šimková, 2012).…”