Scholar
Daryna Dementieva
Google Scholar ID: mLX6olgAAAAJ
TUM
Citations & Impact
Citations
6,969
H-index
16
i10-index
22
Publications
20
Co-authors
68
list available
Publications
Resume (English only)
Academic Achievements
- 1. Paper: Multiverse: Multilingual Evidence for Fake News Detection, MDPI, 2023;
- 2. Paper: ParaDetox: Detoxification with Parallel Data, ACL, 2022;
- 3. Paper: Methods for detoxification of texts for the Russian language, MDPI, 2021;
- 4. Paper: IFAN: An Explainability-Focused Interaction Framework for Humans and NLP Models, submitted to ACL Demo 2023;
- 5. Open Source Contributions: Contributed to multiple open-source projects under s-nlp, including ParaDetox, roberta_toxicity_classifier, bart_detox, etc.
Research Experience
- 1. TUM, Postdoctoral Researcher, Jun 2022 - present, Project: IFAN - An Explainability-Focused Interaction Framework for Humans and NLP Models;
- 2. Skoltech, Research Engineer, Mar 2021 - Mar 2022, Project: Texts Detoxification: Monolingual and Multilingual Setups;
- 3. Beiersdorf, Data Science Intern, Jun 2018 - Aug 2018, Project: News Trend Monitoring for Self-care Novelties Detection;
- 4. Visiology, Data Scientist, Aug 2017 - Aug 2019, Project: NLP2SQL Chatbot for Business Intelligence Platform.
Education
- Ph.D.: Skolkovo Institute of Science and Technology, supervised by Alexander Panchenko, topic: 'Methods for Fighting Harmful Multilingual Textual Content'.
Background
- Research Interests: Applying Large Language Models (in both monolingual and multilingual setups) to different tasks in NLP for social good, such as fake news detection, text style transfer (detoxification), and XAI in NLP. Profile: Currently a postdoctoral researcher at the Social Computing Research Group, Technical University of Munich.
Miscellany
- Personal Interest: Propagating the above-mentioned technologies to the Ukrainian language, as fighting fake news and hate speech is more important than ever for Ukraine.
