Methods for Detoxification of Texts for the Russian Language

TL;DR

This paper explores automatic detoxification methods for Russian texts to reduce offensive language, comparing unsupervised BERT-based and supervised GPT-2-based models, and establishing evaluation metrics.

Contribution

First study to address Russian text detoxification, introducing models and evaluation setup for this language and task.

Findings

Both models effectively reduce offensive content

Unsupervised BERT approach performs local corrections

Supervised GPT-2 approach shows promising results

Abstract

We introduce the first study of automatic detoxification of Russian texts to combat offensive language. Such a kind of textual style transfer can be used, for instance, for processing toxic content in social media. While much work has been done for the English language in this field, it has never been solved for the Russian language yet. We test two types of models - unsupervised approach based on BERT architecture that performs local corrections and supervised approach based on pretrained language GPT-2 model - and compare them with several baselines. In addition, we describe evaluation setup providing training datasets and metrics for automatic evaluation. The results show that the tested approaches can be successfully used for detoxification, although there is room for improvement.