Datasets:
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
Rada_Trees
Overview
Rada_Trees is a comprehensive syntactically annotated corpus of Ukrainian parliament (Verkhovna Rada) transcripts spanning over three decades (1990–2024). This corpus contains approximately 88 million tokens, making it one of the most extensive annotated parliamentary corpora for the Ukrainian language. Paragraphs in Russian and those where Russian words predominate were removed using the CleanText tool.
Key features
- Size: ~88 million tokens
- Period: 1990–2024 (nine convocations)
- Language: Ukrainian
- Source: Official transcripts from plenary sessions downloaded from the Verkhovna Rada open data portal
- Formats: Available in three annotation formats:
Annotation details
The corpus includes comprehensive linguistic annotation.
For data annotated with UD:
- Tokenization and sentence segmentation
- Lemmatization
- Universal POS tags (UPOS) - part-of-speech tagging
- Morphological features (FEATS) - detailed grammatical information
- Dependency relations (HEAD, DEPREL) - syntactic structure
- Format: CoNLL-U (compatible with Universal Dependencies tools and parsers)
For data annotated with nlp_uk:
- Tokenization and sentence segmentation
- Lemmatization
- Morphological tags containing information about pos and morphological features, as well as stylistic annotation (bad, arch, rare, vulg, etc.), according to the nlp_uk standard.
Applications
- Research on Ukrainian morphology and syntax
- Linguistic analysis of parliamentary discourse
- Studies in sociolinguistics and political communication
Technical specifications
- UDPipe2 (ukrainian-parlamint-ud-2.15-241121 model)
- nlp_uk tools based on the LanguageTool engine and VESUM dictionary
Related resources
- ParlaMint-UA: the Ukrainian parliamentary corpus in the ParlaMint framework containing transcripts from 2002 to 2023 (6 convocations)
- GRAC: General Regionally Annotated Corpus of Ukrainian containing transcripts with nlp_uk annotation to the same extent as Rada_Trees, available for online search through NoSketchEngine
- UD_Ukrainian-ParlaMint: a manually annotated Treebank based on Ukrainian parliamentary transcripts that feature 5432 sentences and 84189 tokens
Acknowledgements
We gratefully acknowledge the support of COST Action CA21167 UniDive in the development of the UD Treebank Ukrainian_ParlaMint, which served as the basis for the annotation of this resource.
Licensing
CC BY 4.0
Citation
Lukashevskyi, A., Zakharov, K., & Shvedova, M. (2025). Rada_Trees: A Syntactically Annotated Corpus of Ukrainian Parliament Transcripts (1990-2024).
Availability
The corpus is available for download in all three formats. For access and technical support, please get in touch with the developers.
- Downloads last month
- 47