Resources for Modeling Derivation Using Methods from Distributional Semantics
- Typ
-
ExperimentData
- Autor
-
Sebastian Padó, Aurélie Herbelot, Max Kisselew, Jan Šnajder
-
We are interested in modeling and investigating morphological derivation using methods from Distributional Semantics. Most our work for German is based on DErivBase (Zeller et al., 2013), a derivation lexicon that groups 280K lemmas into 17K derivational families. Other resources used in our recent work can be found below in the "Download" section.
- Referenz
-
Sebastian Padó, Aurélie Herbelot, Max Kisselew, Jan Šnajder:
Predictability of Distributional Semantics in Derivational Word Formation.
In: Proceedings of COLING 2016. Osaka, Japan. - Download
-
Resources from Predictability of Distributional Semantics in Derivational Word Formation:
- Derivationally related word pairs with their features and performance predictions for different CDSMs
(In the partition column "0" corresponds to the training set, "1" to the development set and "2" to the test set)
- Derivationally related word pairs with their features and performance predictions for different CDSMs
Sebastian Padó
Prof. Dr.Lehrstuhlinhaber Theoretische Computerlinguistik, Geschäftführender Direktor des IMS