Tailored Fine-Tuning For The Comma Insertion In Czech

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14210%2F25%3A00142283" target="_blank" >RIV/00216224:14210/25:00142283 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.juls.savba.sk/ediela/jc/2025/1/jc25-01.pdf" target="_blank" >https://www.juls.savba.sk/ediela/jc/2025/1/jc25-01.pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.2478/jazcas-2025-0024" target="_blank" >10.2478/jazcas-2025-0024</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Tailored Fine-Tuning For The Comma Insertion In Czech

Popis výsledku v původním jazyce

Transfer learning techniques, particularly the use of pre-trained Transformers, can be trained on vast amounts of text in a particular language and can be tailored to specific grammar correction tasks, such as automatic punctuation correction. The Czech pre-trained RoBERTa model demonstrates outstanding performance in this task (Machura et al. 2022); however, previous attempts to improve the model have so far led to a slight degradation (Machura et al. 2023). In this paper, we present a more targeted fine-tuning of this model, addressing linguistic phenomena that the base model overlooked. Additionally, we provide a comparison with other models trained on a more diverse dataset beyond just web texts.

Název v anglickém jazyce

Tailored Fine-Tuning For The Comma Insertion In Czech

Popis výsledku anglicky

Transfer learning techniques, particularly the use of pre-trained Transformers, can be trained on vast amounts of text in a particular language and can be tailored to specific grammar correction tasks, such as automatic punctuation correction. The Czech pre-trained RoBERTa model demonstrates outstanding performance in this task (Machura et al. 2022); however, previous attempts to improve the model have so far led to a slight degradation (Machura et al. 2023). In this paper, we present a more targeted fine-tuning of this model, addressing linguistic phenomena that the base model overlooked. Additionally, we provide a comparison with other models trained on a more diverse dataset beyond just web texts.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

60203 - Linguistics

Projekt

—

Návaznosti

R - Projekt Ramcoveho programu EK

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů