This dataset contains 15,019 parallel English-Hausa sentence pairs specifically curated for medical and healthcare translation tasks. It's designed to support machine translation research for the Hausa language in the medical domain.

Dataset Summary

Language pair: English ↔ Hausa
Domain: Medical/Healthcare
Size: 15,019 sentence pairs
Split: 90% train, 10% validation
Generated: 2025-10-02

Supported Tasks

Machine Translation (English to Hausa)
Machine Translation (Hausa to English)
Medical text translation
Cross-lingual medical information retrieval

Dataset Structure

Data Instances

{
  "english": "The patient has a fever.",
  "hausa": "Majinyaci yana da zazzabi."
}

Data Fields

english: English medical/healthcare text
hausa: Corresponding Hausa translation

Data Splits

Train: 13,517 pairs
Validation: 1,502 pairs

Dataset Creation

Source Data

The dataset was generated using advanced natural language processing techniques to create high-quality medical translations covering:

Medical symptoms and conditions (25%)
Medications and treatments (20%)
Doctor-patient consultations (20%)
Prevention and wellness (15%)
Emergency situations (10%)
Medical knowledge base (10%)

Data Quality

All translations were generated using linguistic patterns and medical terminology
Quality validation through automatic checks
Domain-specific vocabulary coverage
Balanced representation across medical subcategories

Uses

Direct Use

This dataset can be used to:

Train machine translation models for English-Hausa medical translation
Fine-tune existing multilingual models for medical domain
Evaluate translation quality in low-resource language settings
Support healthcare accessibility in Hausa-speaking regions

Out-of-Scope Use

General domain translation (optimized for medical content)
Real-time clinical decision making without human oversight
Legal or official medical document translation without verification

Dataset Card Authors

This dataset was created as part of the Hausa Healthcare Translation Project to improve medical accessibility for Hausa-speaking communities.

Licensing Information

This dataset is released under the MIT License.

Citation Information

@dataset{hausa_english_medical_2025,
  title={Hausa-English Medical Translation Dataset},
  author={Healthcare Translation Project},
  year={2025},
  url={https://huggingface.co/datasets/Abduull6771/hausa-english-medical-translation}
}

Contributions

We welcome contributions to improve this dataset. Please see our repository for guidelines on how to contribute.

This dataset is part of efforts to improve healthcare accessibility for Hausa-speaking communities through better language technology.

Downloads last month: 14

Total file size:

48.3 kB