Methods for Detoxification of Texts for the Russian Language

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F21%3A10441629" target="_blank" >RIV/00216208:11320/21:10441629 - isvavai.cz</a>

Výsledek na webu

<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=T6M_T4JBMO" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=T6M_T4JBMO</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.3390/mti5090054" target="_blank" >10.3390/mti5090054</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Methods for Detoxification of Texts for the Russian Language

Popis výsledku v původním jazyce

We introduce the first study of the automatic detoxification of Russian texts to combat offensive language. This kind of textual style transfer can be used for processing toxic content on social media or for eliminating toxicity in automatically generated texts. While much work has been done for the English language in this field, there are no works on detoxification for the Russian language. We suggest two types of models-an approach based on BERT architecture that performs local corrections and a supervised approach based on a pretrained GPT-2 language model. We compare these methods with several baselines. In addition, we provide the training datasets and describe the evaluation setup and metrics for automatic and manual evaluation. The results show that the tested approaches can be successfully used for detoxification, although there is room for improvement.

Název v anglickém jazyce

Methods for Detoxification of Texts for the Russian Language

Popis výsledku anglicky

We introduce the first study of the automatic detoxification of Russian texts to combat offensive language. This kind of textual style transfer can be used for processing toxic content on social media or for eliminating toxicity in automatically generated texts. While much work has been done for the English language in this field, there are no works on detoxification for the Russian language. We suggest two types of models-an approach based on BERT architecture that performs local corrections and a supervised approach based on a pretrained GPT-2 language model. We compare these methods with several baselines. In addition, we provide the training datasets and describe the evaluation setup and metrics for automatic and manual evaluation. The results show that the tested approaches can be successfully used for detoxification, although there is room for improvement.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2021

Kód důvěrnosti údajů

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

Název periodika

Multimodal Technologies and Interaction [online]

ISSN

2414-4088

e-ISSN

—

Svazek periodika

5

Číslo periodika v rámci svazku

9

Stát vydavatele periodika

CH - Švýcarská konfederace

Počet stran výsledku

26

Strana od-do

54

Kód UT WoS článku

000702843200001

EID výsledku v databázi Scopus

2-s2.0-85114864317