Dependency Parsing of Hungarian: Baseline Results and Challenges

Hungarian is a stereotype of morphologically rich and non-configurational languages. Here, we introduce results on dependency parsing of Hungarian that employ a 80K, multi-domain, fully manually annotated corpus, the Szeged Dependency Treebank. We show that the results achieved by state-of-the-art data-driven parsers on Hungarian and English (which is at the other end of the configurational-nonconfigurational spectrum) are quite similar to each other in terms of attachment scores. We reveal the reasons for this and present a systematic and comparative linguistically motivated error analysis on both languages. This analysis highlights that addressing the language-specific phenomena is required for a further remarkable error reduction.