DeriMo 2021

Third Workshop on Resources and Tools
for Derivational Morphology
Nancy, 9-10 September 2021


Building and exploiting resources for derivational morphology: Data-driven and theory-driven approaches

Sebastian Padó
University of Stuttgart

A central characteristic of derivational morphology is its (semi)regularity, i.e., the existence of regular patterns which are however subject to exceptions, gaps and subregularities. In the first part of my talk, I will argue that the creation of derivational resources can profit from the combination of theory-driven and data-driven methods, and will present evidence for this claim from the construction of DErivBase, a derivational dictionary for German, which combines hand written rules, distributional data, and graph theoretic methods [1, 2, 3]. In the second part, I will move to the exploitation of such resources and discuss the tension between how the semantic effects of derivation are captured on the theoretical side (transparency, specificity) and how they are captured on the distributional side [4, 5, 6].


  • [1] DErivBase: Inducing and Evaluating a Derivational Morphology Resource for German. In: Proceedings of ACL, pages 1201-1211. Sofia, Bulgaria, 2013. Britta D. Zeller, Jan Šnajder and Sebastian Padó.
  • [2] Towards Semantic Validation of a Derivational Lexicon. In: Proceedings of COLING, pages 1728-1739. Dublin, Ireland, 2014. Britta Zeller, Sebastian Padó and Jan Šnajder.
  • [3] Evaluating and Improving a Derivational Lexicon with Graph-theoretical Methods. In: Proceedings of DeriMo. Milan, Italy, 2017. Sean Papay, Gabriella Lapesa and Sebastian Padó.
  • [4] Predictability of Distributional Semantics in Derivational Word Formation. In: Proceedings of COLING, pages 1285-1296. Osaka, Japan, 2016. Sebastian Padó, Aurélie Herbelot, Max Kisselew and Jan Šnajder.
  • [5] Are doggies cuter than dogs? Emotional valence and concreteness in German derivational morphology. In: Proceedings of IWCS. Montpellier, France, 2017. Gabriella Lapesa, Sebastian Padó, Tillmann Pross and Antje Rossdeutscher.
  • [6] Grounding Semantic Transparency In Context: A Distributional Semantic Study on German Event Nominalizations. Morphology, 2021. Accepted for publication. Rossella Varvara, Gabriella Lapesa and Sebastian Padó.