The CLASSLA-Stanza model for lemmatisation of non-standard Serbian 2.1

  Authors

Terčon, Luka ; Ljubešić, Nikola and Štefanec, Vanja

  Item identifier

http://hdl.handle.net/11356/1828

 Project URL

https://github.com/clarinsi/classla

 Referenced by

http://dx.doi.org/10.18653/v1/W19-3704

 Date issued

2023-05-10

 Type

toolService

 Language(s)

Serbian

 Description

The model for lemmatisation of non-standard Serbian was built with the CLASSLA-Stanza tool (https://github.com/clarinsi/classla) by training on the SETimes.SR training corpus (http://hdl.handle.net/11356/1200) and the ReLDI-NormTagNER-sr corpus (http://hdl.handle.net/11356/1794), using the srLex inflectional lexicon (http://hdl.handle.net/11356/1233). These corpora were additionally augmented for handling missing diacritics by repeating parts of the corpora with diacritics removed. The estimated F1 of the lemma annotations is ~94.92. The difference to the previous version of the model is that this version is trained on a combination of two corpora (SETimes.SR, ReLDI-NormTagNER-sr).

 Publisher

Jožef Stefan Institute

 Acknowledgement

ARRS (Slovenian Research Agency) P6-0411 "Language Resources and Technologies for Slovene"

Jožef Stefan Institute CLARIN "CLARIN.SI"

ARRS (Slovenian Research Agency) J7-4642 "MEZZANINE"

Connecting Europe Facility (CEF) Telecom INEA/CEF/ICT/A2020/2278341 "MaCoCu - Massive collection and curation of monolingual and bilingual data: focus on under-resourced languages"

 Subject(s)

lemmatisation computer-mediated communication language model

 Collection(s)

CLARIN.SI data & tools

 Other versions

List all versions  

Show full item record