I am a fifth-year Ph.D. Candidate in the Department of Computer Science at the University of Maryland, College Park. I am a member of the CLIP lab working with Marine Carpuat. My research interests are broadly in Multilingual Natural Language Processing (NLP) and Machine Translation.

My recent work focuses on building better models across diverse languages by using humans and AI as joining forces. In my past work, I focused on revisiting common hypotheses adopted when modeling and evaluating multilingual content:

✔️ Improving Machine Translation Through Semantic Analysis: A source text and its translation are not always equivalent in meaning—a common hypothesis made when training machine translation systems. I have worked on building models and algorithms for detecting, analyzing, and mitigating the impact of small cross-lingual meaning differences on machine translation training.

✔️ Revisiting Style Transfer Beyond English: Progress recorded when modeling English does not always port to other languages—a common hypothesis made in Multilingual NLP. I have worked toward creating resources and evaluation models that lay the foundation for studying the task of controlling stylistic variations in natural language generation in a multilingual setting.

Research Interests

  • Computational Semantics
  • Machine Translation
  • Style Transfer
  • Generation Evaluation & Metrics
  • Explainable & Interpretable AI

Industrial Experience

  • Google Research, Summer 2022

    Research Intern

  • Meta AI Research, Summer 2021

    Research Intern

  • Dataminr Research, Summer 2020

    Research Intern

Academic Experience

  • Ph.D. in Computer Science, 2018-now

    University of Maryland, College Park, USA

  • M.Sc. in Computer Science, 2018-2020

    University of Maryland, College Park, USA

  • B.Sc. & M.Eng. in Electrical and Computer Engineering, 2012-2018

    National Technical University of Athens, Athens, Greece

publications

. Searching for Needles in a Haystack: On the Role of Incidental Bilingualism in PaLM's Translation Capability. In Proceedings of the 61th Annual Meeting of the Association for Computational Linguistics (ACL), 2023.

pdf cite

. Understanding and Detecting Hallucinations in Neural Machine Translation via Model Introspection. In Transactions of the Association for Computational Linguistics (TACL), 2023.

pdf cite

. BitextEdit: Automatic Bitext Editing for Improved Low-Resource Machine Translation. In Proceedings of the Findings of the 2022 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL-Findings), 2022.

pdf slides poster talk cite

. Can Synthetic Translations Improve Bitext Quality?. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (ACL), 2022.

pdf data slides poster talk cite

. Evaluating the Evaluation Metrics for Style Transfer: A Case Study in Multilingual Formality Transfer. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language (EMNLP) Processing, 2021.

pdf code slides talk cite

projects

.js-id-All

Bitext Refinement

Mined bitexts can contain imperfect translations that yield unreliable training signals for Neural Machine Translation. While filtering …

Detecting Semantic Divergences At Scale

Quantifying fine-grained cross-lingual semantic divergences at scale, requires computational models that do not rely on human-labeled …

Divergences in Machine Translation

Parallel texts—a source paired with its (human) translation—are routinely used for training machine translation systems …

Multilingual Evaluation

As a community, we have overfitted the characteristics of English-language data when modeling various tasks, does the same hold for our …

Multilingual Style Transfer

A dominant hypothesis in multilingual research is that models developed and optimized for English can be seamlessly transferred (and …

Rationalized Semantic Divergences

Detecting fine-grained semantic divergences—small meaning differences in segments that are treated as exact translation …

contact