I work on how language models learn robust, useful representations under real constraints — limited data, limited compute, shifting domains. I'm currently building gake.
PhD in NLP, Aristotle University of Thessaloniki (2021, Intelligent Systems Lab, adv. Ioannis Vlahavas). MSc Big Data & Text Analytics and BSc CS, University of Essex (w/ Udo Kruschwitz, Massimo Poesio). Previously postdoc at CERTH-ITI / M4D.
News
- 2025
- Paper on building open-retrieval conversational QA systems with synthetic data accepted at SIGDIAL 2025.
- 2025
- Paper on explainable text classification pipeline accepted at MDTT 2025
- 2024
- Joined Omilia — Conversational Intelligence as a NLP Research Scientist
- 2024
- Paper on cross-domain hate speech detection accepted at IEEE/WIC WI-IAT 2024.
- 2022
- Acting Adjunct Lecturer at the AUTh School of Informatics.
Selected publications
- Building Open-Retrieval Conversational Question Answering Systems by Generating Synthetic Data and Decontextualizing User Questions.2025 · SIGDIAL
- A Multi-Task Text Classification Pipeline with Natural Language Explanations: A User-Centric Evaluation in Sentiment Analysis and Offensive Language Identification in Greek Tweets.2025 · MDTT (CEUR Workshop Proceedings)
- Cross-Domain Hate Speech Detection for Content Moderation in Greek Social Networks.2024 · IEEE/WIC WI-IAT
- Domain-Aligned Data Augmentation for Low-Resource and Imbalanced Text Classification.2023 · ECIR
- CoreLM: Coreference-aware Language Model Fine-Tuning.2021 · CRAC @ EMNLP
- TransforMED: End-to-end Transformers for Evidence-based Medicine and Argument Mining in medical literature.2021 · Journal of Biomedical Informatics