TatarNLPWorld/tatarstan-toponyms

Dataset Card for Toponyms of Tatarstan

Dataset Details

Dataset Description

A comprehensive dataset of 9,688 toponyms (place names) from Tatarstan and Tatar-populated regions, curated by TatarNLPWorld as part of the Tat2Vec project. Each entry provides detailed linguistic, geographical, and etymological information about Tatar and Russian place names. The dataset is specifically designed for linguistic research, onomastic studies, and training NLP models for named entity recognition and geocoding tasks.

• Curated by: TatarNLPWorld team (Mullosharaf K. Arabov and contributors)
• Language(s) (NLP): Tatar (tt), Russian (ru)
• License: other – see Licensing & Legal Notice below.

Licensing & Legal Notice

This dataset follows the practice established by large web-crawled corpora such as HPLT and OSCAR:

• Original source texts (e.g., etymological explanations, bibliographic citations) are not owned by the TatarNLPWorld team and are not covered by the CC BY-SA 4.0 license.
• The structured compilation, metadata, and annotations created by TatarNLPWorld are licensed under CC BY-SA 4.0.
• Users are solely responsible for ensuring their use of the underlying texts complies with applicable copyright law. For commercial use of verbatim excerpts, permission from the original copyright holders may be required.
• A notice-and-takedown procedure is in place: rights holders can request removal of specific entries by contacting the dataset maintainers (see Dataset Card Contact). We commit to responding within 14 business days and removing disputed content in the next release.

Dataset Sources

• Repository: https://huggingface.co/datasets/TatarNLPWorld/tatarstan-toponyms
• Paper: arXiv:2605.05962
  “Tatarstan Toponyms: A Bilingual Dataset and Hybrid RAG System for Geospatial Question Answering” (Computational Linguistics in Bulgaria, 2026)
• Demo: https://huggingface.co/spaces/TatarNLPWorld/tatarstan-toponyms-qa

Uses

Direct Use

The dataset is intended for research and development in:

• Named Entity Recognition (NER) for Tatar and Russian geographic names
• Token classification and sequence labelling
• Geocoding and reverse geocoding
• Question answering over geospatial knowledge
• Linguistic and onomastic studies (etymology, toponym classification)
• Cross-lingual alignment of place names
• Feature extraction for geographic terminology

Out-of-Scope Use

• Commercial redistribution of verbatim source excerpts (e.g., etymological descriptions) without obtaining rights from the original publishers.
• Any use that violates the copyright of the underlying dictionaries, encyclopedias, or academic works.
• Fully automated decision-making that could have legal or safety implications without human review.
• Use as a sole source for geographic navigation — coordinates may contain errors or refer to approximate locations.

Dataset Structure

The dataset is provided in two splits:

Split	Size	Description
full	9,688	Complete dataset with all records
sample	100	Small random sample for testing and exploration

Field Descriptions

Each record contains the following columns (names are in Russian, as in the original source):

Field	Type	Description
id	int64	Unique identifier from source
url	string	Source URL with additional information
тип_топонима	string	Type category: Топоним / Микротопоним
вид_топонима	string	Kind category: Ойконим, Гидроним, Ороним, Без типа
географический_объект	string	Geographic object type (76 unique values)
на_русском	string	Place name in Russian (Cyrillic)
на_татарском	string	Place name in Tatar (Cyrillic)
субъект_федерации	string	Federal subject (region)
физико_географические_сведения	string	Physical-geographic characteristics
географическое_положение	string	Geographic location description
этимология_названия	string	Name etymology and origin (verbatim excerpts from copyrighted sources)
источники	string	Bibliographic sources (full citations of original works)
latitude	float64	Decimal latitude (when available)
longitude	float64	Decimal longitude (when available)
has_map	bool	Whether a map is available on the source page

Dataset Creation

Curation Rationale

The dataset was created to fill a gap in openly available, high-quality multilingual resources for Tatar toponymy. It supports NLP tasks such as named entity recognition, geocoding, and cross-lingual mapping, as well as linguistic research on the etymology and structure of place names. The resource is part of the larger Tat2Vec project aimed at advancing Tatar language technology.

Source Data

Data Collection and Processing

Data were extracted from the digital archive toponym.antat.ru, which itself aggregates information from authoritative printed sources (dictionaries, encyclopedias, academic monographs). The raw HTML pages were parsed, and fields such as name, coordinates, etymology, and bibliographic references were extracted and structured into a tabular format. Geolocation data (latitude/longitude) were obtained from the source pages where available; missing coordinates were left empty. No automatic translation or paraphrasing was applied to textual fields; they preserve the original wording of the cited works.

Who are the source data producers?

The original data were produced by scholars and institutions specializing in Tatar onomastics, including:

• Ф.Г. Гарипова (F.G. Garipova) – hydronym dictionaries
• Г.Ф. Саттаров (G.F. Sattarov) – Tatar toponymy
• Р.Г. Әхмәтьянов (R.G. Akhmetyanov) – etymological dictionary
• Татарская энциклопедия (Tatar Encyclopedia, Institute of Tatar Encyclopedia, Academy of Sciences of the Republic of Tatarstan)
• Other authors cited in the источники field

The digital compilation at toponym.antat.ru was used as the primary access point, but the intellectual content belongs to the original authors and publishers.

Annotations

The dataset does not contain new human annotations in the traditional sense; rather, it is a structured extraction of existing information. However, the TatarNLPWorld team performed:

• Parsing and normalization of the source HTML
• Extraction of geographic coordinates and mapping flags
• Assignment of unique IDs
• Creation of the full and sample splits

No crowdsourced or manual annotation was added beyond the original source content.

Personal and Sensitive Information

The dataset contains no personal data about individuals. It consists solely of geographic names, their descriptions, and bibliographic references. No anonymization was required.

Bias, Risks, and Limitations

• Geographic bias: The dataset focuses overwhelmingly on the Republic of Tatarstan (~93%) and may not represent Tatar toponyms in other regions equally.
• Coverage gaps: Approximately 7% of records lack coordinates; some toponyms may be missing or have outdated names.
• Source reliability: The textual fields (этимология_названия, источники) are taken verbatim from academic sources and may reflect the historical or theoretical positions of their authors. They should be used with awareness of potential outdated or contested interpretations.
• Copyright constraints: The dataset includes copyrighted excerpts. Users must exercise caution when redistributing or using these excerpts beyond fair use/citation.
• Language imbalance: Although bilingual, some records may have missing translations in one of the languages.

Recommendations

• For NLP training, consider filtering records with missing coordinates or incomplete etymology.
• When using the dataset for commercial applications, verify the copyright status of any verbatim text you plan to reuse.
• Always cite the original sources in addition to the dataset itself.
• If you are a rights holder and find your work included, please contact us for removal.

Citation

If you use this dataset, please cite both the dataset and the accompanying paper.

BibTeX (dataset):

@dataset{tatarstan_toponyms_2026,
    title = {Toponyms of Tatarstan Dataset},
    author = {Arabov, Mullosharaf Kurbonovich},
    year = {2026},
    publisher = {Hugging Face},
    version = {1.0.0},
    url = {https://huggingface.co/datasets/TatarNLPWorld/tatarstan-toponyms}
}

BibTeX (paper):

@article{arabov2026tatarstan,
    title = {Tatarstan Toponyms: A Bilingual Dataset and Hybrid RAG System for Geospatial Question Answering},
    author = {Arabov, Mullosharaf K. and Khaybullina, Svetlana S. and Naumetova, Dinara M.},
    year = {2026},
    journal = {Computational Linguistics in Bulgaria},
    volume = {2},
    number = {1},
    pages = {41--63},
    url = {https://arxiv.org/abs/2605.05962},
    doi = {10.47810/JCLIB.2.2026.03}
}

APA (paper):
Arabov, M. K., Khaybullina, S. S., & Naumetova, D. M. (2026). Tatarstan Toponyms: A Bilingual Dataset and Hybrid RAG System for Geospatial Question Answering. Computational Linguistics in Bulgaria, 2(1), 41–63. https://doi.org/10.47810/JCLIB.2.2026.03

Glossary

• Топоним (Toponym) – a proper name of a geographic feature.
• Микротопоним (Microtoponym) – a name of a small local feature (field, spring, etc.).
• Ойконим (Oikonym) – a name of a settlement (village, town, city).
• Гидроним (Hydronym) – a name of a water body (river, lake, spring).
• Ороним (Oronym) – a name of a relief feature (mountain, hill, valley).
• Этимология (Etymology) – the study of the origin and historical development of a word.

More Information

• The dataset is part of the Tat2Vec project.
• A hybrid geospatial question-answering system based on this dataset is described in the accompanying paper.
• The dataset is provided “as is” without warranty of any kind.

Dataset Card Authors

• Mullosharaf K. Arabov (TatarNLPWorld)
• Contributions from the TatarNLPWorld community.

Dataset Card Contact

For questions, takedown requests, or collaboration, please contact:
📧 marabov@kpfu.ru