VytautoDidziojoUniversitetas/LT_Morphosyntax_Corpus_SIMAS

Morphologically and Syntactically Annotated Corpus SIMAS

Description

SIMAS belongs to the category of general annotated corpora of written language. The corpus consists of texts written and published between 2000 and 2025 and represents original contemporary standard written Lithuanian. The corpus is composed of four sections reflecting different registers: fiction, academic writing, administrative texts, and journalism (texts published on the Lithuanian National Radio and Television portal LRT.lt).

Annotation

The corpus was automatically annotated morphologically and syntactically and subsequently reviewed by linguists. Automatic morphological annotation was carried out using the tool “Morfuoklis“ (https://sitti.vdu.lt/morfuoklis/lt). Automatic syntactic parsing was performed using the international tool UDPipe (https://lindat.mff.cuni.cz/services/udpipe/) adapted for the Lithuanian language.

Annotation standards

The corpus is annotated according to the international Universal Dependencies (UD) standard (see Universal Dependencies for Lithuanian, https://universaldependencies.org/lt/).

The following standards were also used for morphological analysis: Jablonskis morphological annotation standard (Rimkutė E., Bielinskienė A., Boizou L., Utka A., & Dadurkevičius V. (2024)). (https://hdl.handle.net/20.500.11821/56); MULTEXT-East (http://corpus.vdu.lt/lt/morph).

For syntactic analysis, the Universal Dependencies Standard: the Lithuanian Syntactic Annotation Guidelines (Rimkutė, E., Bielinskienė, A., Boizou, L., Kamandulytė-Merfeldienė, L., Brokaitė, K., Jancaitė-Skarbalė, L., Stepšys, J., & Žemrietė, M. (2026)), https://sitti.vdu.lt/wp-content/uploads/2026/05/UD_standarto_gaires.pdf was applied.

Format

The corpus is distributed as a single file. The file is provided in CoNLL-U format. CoNLL-U is a text-based data format used to annotate natural language texts according to the UD standard. Each line in a file represents a single word (token), with tab-separated columns providing detailed information about its lemma, part of speech, morphological features, and syntactic relations to other words in the sentence.

Data composition

The proportions are as follows: journalism (multi-topic and multi-genre articles from LRT.lt) comprises 4,852,633 words (approximately 50% of the corpus). Administrative documents account for 2,085,581 words, while academic texts (books and articles from various scientific fields) contain 2,108,574 words (each representing about 20% of the corpus). Fiction texts comprise 963,616 words (about 10% of the corpus).

Distribution of text types

Text type	Token count	Word count	Sentence count
Administrative literature	2,413,788	2,085,581	104,811
Journalism / Publicistic writing	5,977,225	4,852,633	342,876
Scientific literature (books)	1,415,277	1,163,045	70,119
Scientific literature (articles)	1,169,008	945,529	51,006
Fiction literature	1,246,277	963,632	100,157
Total	12,221,575	10,010,420	668,969

Licensing

This dataset is distributed under the NewGenLTU OpenRAIL-D licence.

This licence was specifically created by a team of researchers at Vytautas Magnus University for the publication of datasets produced in the course of research and development projects funded by the New Generation Lithuania plan. The licence aims to support open and responsible downstream use while ensuring compliance with applicable requirements related to copyright, personal data protection, and the EU AI Act.

Please see the accompanying LICENSE.txt file in this repository for the full licence text.

Intended use

This dataset may be used for a range of Lithuanian-language NLP and AI tasks, including:

• tokenization,
• token classification,
• text classification,
• grammar and style correction,
• text analysis,
• other language technology applications.

Restrictions

Under the NewGenLTU OpenRAIL-D licence, users must not use the dataset for:

• discrimination,
• the development of weapons or military applications,
• automated decision-making affecting people,
• disinformation,
• privacy infringement,
• medical or health advice,
• creation of malware,
• harassment,
• dishonest research,
• collection, extraction, or reconstruction of personal data.

Please refer to the accompanying LICENSE.txt file for the full and authoritative wording.

Citation

Please cite the dataset as:

Vytautas Magnus University. 2026. Abstract Corpora for Artificial Intelligence. Hugging Face. https://huggingface.co/datasets/VytautoDidziojoUniversitetas/LT_Morphosyntax_Corpus_SIMAS

BibTeX:

@dataset{lt_morphosyntax_corpus_simas_2026,
  author    = {Vytautas Magnus University},
  title     = {Morphologically and Syntactically Annotated Corpus SIMAS},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/VytautoDidziojoUniversitetas/LT_Morphosyntax_Corpus_SIMAS},
  note      = {Developed under the New Generation Lithuania plan, project 02-098-K-0001.}
}

Acknowledgement

Funded by the European Union — NextGenerationEU. Naujos kartos Lietuva.

This project is funded by the Recovery and Resilience Facility under the Lithuanian "Naujos kartos Lietuva" plan and by the State Budget of the Republic of Lithuania.

Project: Morphologically and syntactically annotated text models for training (gold standards) — Project No. No. 02-098-K-0001.