Skoltech/russian-inappropriate-messages

The concept of inappropriateness is described in this article presented at the workshop for Balto-Slavic NLP at the EACL-2021 conference . The model was trained, validated and tested only on the samples with 100% confidence . This article describes the first version of the dataset, while the model is trained on the extended version . The properties of the data are the same as the one described in the article, the only difference is the size. The model is open-sourced on our GitHub or on kaggle. Please note that this article does not describe the model as a model for the extended dataset. The extended dataset is the same thing described in article,