Syntax is the Key to Semantics: Universal Dependencies and Abstract Meaning Representation

This dataset is an AMR annotation of the 1000 sentences of the English-Parallel-UD treebank provided by the Universal Dependencies (UD) project. In addition to English, UD provides the same 1000 sentences translated by professional translators into 21 languages:

• Arabic, Chinese, Czech, Finnish, French, Galician, German, Hindi, Indonesian, Icelandic, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Thai, and Turkish.

750 of the sentence are originally in English (IDs starting with n01 or w01), the other 250 sentences are originally in German (02), French (03), Italian (04) or Spanish (05) and translated via English to the other languages.

For example the following graph represents the sentence n01001013:

And corresponds to the dependency syntax trees in different languages:

• English
• French
• Turkish
• Arabic
• Finnish
• German
• Hindi
• Chinese

Licence

• The sentences from UD-English-PUD are licensed under the CC BY-SA 3.0
• The AMR graphs are licensed under Attribution-ShareAlike 4.0 International, CC BY-SA 4.0

Reference

If you use this dataset, please cite our article:

@inproceedings{heinecke2026-PUD,
  author = {Heinecke, Johannes},
  booktitle = {Ninth Workshop on Universal Dependencies @ LREC 2026},
  title = {{Syntax is the Key to Semantics: Universal Dependencies and Abstract Meaning Representation}},
  year = {2026},
  address = {Palma de Mallorca, Spain}
}