Hello!
My name is Nikolay Babakov. I am a Ph.D. student at the University of Santiago de Compostela.
I am interested in NLP and explainable AI. My current research is Natural Language Generation for Bayesian Networks reasoning explanation. My recent research was dedicated to toxicity detection and text style transfer.
You may reach me in telegram @bbkjunior or via email: [email protected] or [email protected]
ButterflAI effect - the blog where I collect interesting examples of application of AI to the real tasks
- Russian Sensitive Topics Classifier - model for detecting flammabel/senstitve topics (e.g. racism, drugs, terrorism, etc.)
- Russian Inappropriate Utterances Classifier - model for detecting nontoxic but still inappropriate messages
- Mutual Implication Score - symmetric measure of text semantic similarity based on a RoBERTA model pretrained for natural language inference and fine-tuned on a paraphrase detection dataset
- GenChal_2022_nigula - the model gets the English text with the error and the exact span of the error and should return the comment in natural language, which explains the nature of the error.
- LEWIP-informal - model designed to transfer formal text into informal keeping the important slots from the source text. The slots can be either pre-defined or detected automatically.
- mdistilbert-formality-ranker - multilingual classifier of textual formality
- deberta-large-formality-ranker - English classifier of textual formality
- Second place (team nigula) at GenChall2022 - task dedicated to generating feedback to the English phrases containing errors.