Weird Inflects but OK: Making Sense of Morphological Generation Errors
Kyle Gorman, Arya McCarthy, Ryan Cotterell, Miikka Silfverberg, Ekaterina Vylomova, Magdalena Markowska
September 2019

We conduct a manual error analysis of the CoNLL-SIGMORPHON 2017 Shared Task on Morphological Reinflection. In this task, systems are given a word in citation form (e.g., hug) and asked to produce the corresponding inflected form (e.g., the simple past hugged). This design lets us analyze errors much like we might analyze children's production errors. We propose an error taxonomy and use it to annotate errors made by the top two systems across twelve languages. Many of the observed errors are related to inflectional patterns sensitive to inherent linguistic properties such as animacy or affect; many others are failures to predict truly unpredictable inflectional behaviors. We also find nearly one quarter of the residual "errors" reflect errors in the gold data.
Format: [ pdf ]
Reference: lingbuzz/004787
(please use that when you cite this article)
Published in: To appear in CoNLL '19
keywords: phonology, morphology, allomorphy, nlp, inflection, generation, error analysis, morphology, phonology
previous versions: v1 [September 2019]
Downloaded:209 times


[ edit this article | back to article list ]