The SLANGUAGE Tiv Parallel Corpus is the first commercially developed parallel English–Tiv text dataset, created by STECH Global Limited through the SAVA Languages (SLANGUAGE) project. Tiv is a tonal Bantoid language spoken by approximately 8 million people, primarily in Benue State, Nigeria. Despite its significant speaker population, Tiv remains severely underrepresented in AI and NLP research globally.

This release (v1) contains 1,169 verified parallel sentences across four domains, created and reviewed by native Tiv speakers with strong linguistic backgrounds to ensure tonal accuracy and natural language use.

Dataset Details

Category	Sentences
Agriculture	200
Daily Life	200
Religion	378
Common Questions	391
Total	1,169

Language Information

Language: Tiv (ISO 639-3: tiv)
Language family: Niger-Congo, Bantoid
Region: Benue State, Nigeria
Speakers: ~8 million
Script: Latin
Tonal: Yes — two primary tones (high and low) that are phonemically distinctive

Data Collection

Sentences were written by the project director and reviewed by a team of four native Tiv speakers with university-level education to ensure linguistic accuracy, tonal correctness and natural language use. This is not a web-scraped or automatically generated dataset. Every sentence has been manually verified.

Format

Each file is a CSV with three columns:

id — unique sentence identifier
english — English source sentence
tiv — Tiv translation

Intended Use

This dataset is intended for:

NLP and machine learning research
Tiv language technology development
Academic and linguistic study
Humanitarian and non-commercial applications

This dataset is not licensed for commercial use. See licence terms below.

Limitations

This is v1 — a partial release of a larger verified corpus already completed and awaiting staged release
Audio recordings are in progress and will be released separately
Additional categories will be added in future releases
A second linguistic team is currently producing an independent Tiv translation of the full corpus to capture dialectal variation across Tiv-speaking regions

Future Releases

STECH Global Limited is building a full Tiv voice dataset including text, audio recordings, and trained TTS and ASR models. Follow this repository for updates.

Licence

Creative Commons Attribution Non-Commercial 4.0 International (CC BY-NC 4.0)

You are free to use, share and adapt this dataset for non-commercial purposes, provided you give appropriate credit to STECH Global Limited.

Citation

If you use this dataset in your research or project, please cite:

@dataset{slanguage_tiv_2026,
  title     = {SLANGUAGE Tiv Parallel Corpus v1},
  author    = {Unenge, Sooter},
  year      = {2026},
  publisher = {STECH Global Limited},
  url       = {https://huggingface.co/datasets/SLANGUAGE/SLANGUAGE_TIV_001}
}

Contact

STECH Global Limited savasamuel217@gmail.com

Downloads last month: 25

Total file size:

121 kB