Daniel Zeman

Main Research Interests

My research in computational linguistics is mostly centered around morphological analysis and dependency syntax of natural languages. I am interested in computational models of morphology and syntax, including preparation of annotated data to train such models.

I am particularly interested in multilingual approaches that work for many typologically different languages, cross-lingual techniques that support disadvantaged languages with little resources, and harmonized data resources that support multilingual processing. Besides training models, I also look into ways how these datasets can be used in comparative linguistics and typology.

Beyond surface syntax I am working on deep syntactic relations, semantic roles and coreference. In the past I also worked on statistical machine translation for morphologically rich languages.

Selected Bibliography

• Google Scholar
• ORCID: 0000-0002-5791-6568
• Scopus ID: 23092520600
• Researcher ID: B-2844-2009

1. Marie-Catherine de Marneffe, Christopher Manning, Joakim Nivre, Daniel Zeman (2021): Universal Dependencies. In: Computational Linguistics, ISSN 1530-9312, vol. 47, no. 2, pp. 255-308 (url, local PDF, bibtex)
2. Daniel Zeman, Jan Hajič (2020): FGD at MRP 2020: Prague Tectogrammatical Graphs. In: Proceedings of the CoNLL 2020 Shared Task: Cross-Framework Meaning Representation Parsing, pp. 33-39, Association for Computational Linguistics, Stroudsburg, PA, USA, ISBN 978-1-952148-64-4 (url, local PDF, bibtex)
3. Zdeněk Žabokrtský, Daniel Zeman, Magda Ševčíková (2020): Sentence Meaning Representations across Languages: What Can We Learn from Existing Frameworks? In: Computational Linguistics, ISSN 1530-9312, vol. 46, no. 3, pp. 605-665 (url, local PDF, bibtex)
4. Daniel Zeman (2018): The World of Tokens, Tags and Trees. ISBN 978-80-88132-09-7 (url, bibtex)
5. Daniel Zeman, Jan Hajič, Martin Popel, Martin Potthast, Milan Straka, Filip Ginter, Joakim Nivre, Slav Petrov (2018): CoNLL 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. In: Proceedings of the CoNLL 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies, pp. 1-21, Association for Computational Linguistics, Stroudsburg, PA, USA, ISBN 978-1-948087-82-7 (pdf, local PDF, bibtex)
6. Daniel Zeman, Ondřej Dušek, David Mareček, Martin Popel, Loganathan Ramasamy, Jan Štěpánek, Zdeněk Žabokrtský, Jan Hajič (2014): HamleDT: Harmonized Multi-Language Dependency Treebank. In: Language Resources and Evaluation, ISSN 1574-020X, vol. 48, no. 4, pp. 601-637 (url, local PDF, bibtex)
7. Martin Popel, David Mareček, Jan Štěpánek, Daniel Zeman, Zdeněk Žabokrtský (2013): Coordination Structures in Dependency Treebanks. In: Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pp. 517-527, Association for Computational Linguistics, Sofija, Bulgaria, ISBN 978-1-937284-50-3 (pdf, local PDF, poster, slides, bibtex)
8. Daniel Zeman (2008): Reusable Tagset Conversion Using Tagset Drivers. In: Proceedings of the 6th International Conference on Language Resources and Evaluation (LREC 2008), pp. 213-218, European Language Resources Association, Marrakech, Morocco, ISBN 2-9517408-4-0 (url, poster, local PDF, bibtex)
9. Daniel Zeman, Philip Resnik (2008): Cross-Language Parser Adaptation between Related Languages. In: IJCNLP 2008 Workshop on NLP for Less Privileged Languages, pp. 35-42, International Institute of Information Technology, Hyderabad, India (url, local PDF, slides, bibtex)

See here for the complete list of publications.

Habilitation slides.