TajikMatchingQA is a matching question answering dataset in the Tajik language, containing 16,747 matching groups across 571 subjects spanning 9 knowledge categories. Each group consists of an instruction, a list of items, a list of options, and correct pairs mapping items to options. The dataset is designed for training and evaluating models on matching tasks in low-resource languages, with a special focus on Central Asian linguistic diversity.

Curated by: TajikNLPWorld
Language(s) (NLP): Tajik (tg)
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 question texts (instructions, item lists, option lists) remain the property of their respective authors and educators. They are not owned by the TajikNLPWorld team and are not covered by the CC BY‑SA 4.0 license.
The structured compilation, metadata, annotations (correct pair mappings), and packaging created by TajikNLPWorld are released under the CC BY‑SA 4.0 license.
Users are solely responsible for ensuring their use of the underlying content 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 content 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/TajikNLPWorld/TajikMatchingQA

Uses

Direct Use

The dataset is intended for:

Training and evaluating matching (pair‑matching) models for the Tajik language.
Benchmarking low‑resource language models on structured knowledge tasks.
Research on cross‑lingual transfer and domain adaptation for Central Asian languages.
Educational applications such as automated quiz generation, assessment tools, and interactive learning systems.

Out-of-Scope Use

The dataset should not be used:

As a source of truth for critical decisions (e.g., medical, legal, or safety‑critical systems).
For any purpose that violates the original copyright of the underlying texts (see Licensing & Legal Notice).
To discriminate against or misrepresent any group of people.

Dataset Structure

The dataset consists of a single split (train) with 16,747 examples. Each example is a matching group containing an instruction, two lists (items and options), and the correct pairings between them.

Data Fields

Field	Type	Description
`group_id`	string	Unique group identifier
`instruction`	string	Instruction text (e.g., "Динҳо ва номи ибодатгоҳҳои онҳоро муайян намоед:")
`subject`	string	Academic subject name (in Tajik)
`category`	string	Knowledge domain category (one of 9 categories)
`items`	list[string]	List of items to be matched
`options`	list[string]	List of possible options
`correct_pairs`	list[list[int]]	Pairs of `(item_index, option_index)` indicating correct matches

Example Record

Field	Value
`group_id`	`1`
`instruction`	`Динҳо ва номи ибодатгоҳҳои онҳоро муайян намоед:`
`subject`	`Диншиносӣ`
`category`	`Humanities`
`items`	`["Ислом", "Масеҳият", "Яҳудият"]`
`options`	`["Масҷид", "Калисо", "Синанога"]`
`correct_pairs`	`[[0, 0], [1, 1], [2, 2]]`

Knowledge Categories

Category	Groups	Percentage
Humanities	5,256	31.4%
Other	3,840	22.9%
Natural Sciences	2,960	17.7%
Economics & Business	2,006	12.0%
Engineering & Technology	1,892	11.3%
Sports & Tourism	442	2.6%
Medicine & Healthcare	200	1.2%
Agriculture	111	0.7%
Military Science	40	0.2%

Dataset Creation

Curation Rationale

Tajik, a low‑resource language of the Persian family, lacks large‑scale structured benchmarks for knowledge‑intensive tasks beyond simple multiple‑choice. Matching tasks require a deeper understanding of relationships between concepts, which is valuable for educational and reasoning models. TajikMatchingQA was created to fill this gap by providing a diverse, high‑quality matching dataset that can serve as a foundation for training and evaluating such models in Tajik, as well as for studying cross‑lingual transfer from higher‑resource languages.

Source Data

Data Collection and Processing

The dataset was collected by the authors during their work at the Tajik National University from 2012 to 2016, and subsequently expanded through collaborations with colleagues from the following institutions:

Tajik National University
Tajik State Pedagogical University named after Sadriddin Aini
Bokhtar State University named after Nosir Khusrav
Other higher education institutions of Tajikistan

The collection process continued over the period 2012–2026. Matching groups were compiled from various academic sources, examination materials, and educational resources in the Tajik language. Data processing included:

Filtering out duplicates and low‑quality groups.
Standardizing the format (one instruction, two lists, correct pairs).
Assigning each group to a subject and a high‑level knowledge category.
Validating that every item has a unique correct option and that the number of items equals the number of correct pairs.

No automated translation was used; all content is originally in Tajik.

Who are the source data producers?

The source data producers are educators, lecturers, and content creators from the Tajik universities listed above. The matching groups reflect the curricula and examination standards of Tajik higher education institutions during the collection period.

Annotations

Annotation process

Annotations (correct pairings) were provided by the same educators who authored or curated the matching groups. No separate annotation campaign was conducted.

Who are the annotators?

The annotators are the same individuals who created the matching groups — subject matter experts from Tajik universities.

Personal and Sensitive Information

The dataset consists of general knowledge and educational matching exercises. It does not contain personal data, addresses, names, or any information that could be used to identify individuals. No anonymization was necessary.

Bias, Risks, and Limitations

Over‑representation of certain subjects or domains (e.g., Humanities).
Under‑representation of rapidly evolving fields.
Possible factual inaccuracies or outdated information due to the long collection period.
Limited to the Tajik language; may not generalize to other Persian dialects.

Recommendations

Be aware of the domain distribution when evaluating models.
Avoid using the dataset for high‑stakes decisions without additional validation.
Respect the copyright of individual texts (see Licensing & Legal Notice).
Consider combining with other resources for a more balanced evaluation.

Citation

BibTeX (dataset):

@misc{tajikmatchingqa2026,
  author = {TajikNLPWorld},
  title = {TajikMatchingQA: A Matching Question Answering Dataset for Tajik},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/TajikNLPWorld/TajikMatchingQA}
}

APA (dataset): TajikNLPWorld. (2026). TajikMatchingQA: A Matching Question Answering Dataset for Tajik [Data set]. Hugging Face. https://huggingface.co/datasets/TajikNLPWorld/TajikMatchingQA

Glossary

Matching QA – a task where a model must correctly pair items from one list with options from another, based on an instruction.
Low‑resource language – a language with limited digital resources for NLP.
Central Asian languages – languages spoken in Central Asia, including Tajik, Uzbek, Kazakh, etc.

More Information

For additional details, updates, or to report issues, please visit the Hugging Face dataset repository.

Dataset Card Authors

TajikNLPWorld (community)

Dataset Card Contact

For questions, takedown requests, or collaboration, please open an issue on the Hugging Face repository:
https://huggingface.co/datasets/TajikNLPWorld/TajikMatchingQA/discussions
📧 Direct email: cool.araby@gmail.com

Downloads last month: 14

Total file size:

2.73 MB