Ghana-QA: Ghanaian Question-Answer Pairs Dataset

Dataset Summary

The Ghana-QA dataset consists of question-answer pairs derived from Ghanaian news articles. It contains questions in three widely spoken Ghanaian languages and their corresponding answers.

• Total QA Pairs: 3,547,660
• Languages: Twi, Ewe, Ga
• Source: Ghanaian news articles
• Project: Part of the GhanaNLP initiative

Data Fields

• question: The question text
• answer: The corresponding answer text
• lang: Language code (twi/ewe/ga)

How to Use

The dataset is provided as a single CSV file with all languages combined and shuffled.

import pandas as pd

# Load the dataset
df = pd.read_csv("dataset.csv")

# View sample data
print(df.head())
print(f"Total QA pairs: {len(df)}")

# Filter by language
twi_qa = df[df['lang'] == 'twi']
ewe_qa = df[df['lang'] == 'ewe']
ga_qa = df[df['lang'] == 'ga']

print(f"Twi QA pairs: {len(twi_qa)}")
print(f"Ewe QA pairs: {len(ewe_qa)}")
print(f"Ga QA pairs: {len(ga_qa)}")

Loading with Hugging Face Datasets

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("ghananlpcommunity/ghana-qa")

# Access the data
print(dataset['train'][0])

# Filter by language
twi_data = dataset['train'].filter(lambda x: x['lang'] == 'twi')

Language Distribution

The dataset includes questions and answers in three Ghanaian languages:

• Twi (twi): Akan language spoken in Ghana
• Ewe (ewe): Niger-Congo language spoken in Ghana and Togo
• Ga (ga): Kwa language spoken primarily in the Greater Accra Region

Dataset Creation

This dataset was created by extracting question-answer pairs from Ghanaian news articles, providing a valuable resource for:

• Question-answering model training
• Multilingual NLP research
• Low-resource language model development
• Cross-lingual transfer learning

Citation Information

If you use this dataset, please cite:

@dataset{Ghana-QA_2025,
  title = {Ghana-QA: Ghanaian Question-Answer Pairs},
  author = {Mich-Seth Owusu and Priscilla Lartey and Gerhardt Datsomor},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/ghananlpcommunity/ghana-qa}
}

License

This dataset is licensed under CC-BY-NC-4.0 (Creative Commons Attribution-NonCommercial 4.0 International).

Acknowledgments

Special thanks to the primary contributors of the Ghana-QA project: Mich-Seth Owusu, Priscilla Lartey, Gerhardt Datsomor, and all other volunteers from the GhanaNLP community. This work is part of the GhanaNLP initiative to advance natural language processing for Ghanaian languages.

Contact

For questions or feedback, please reach out through the GhanaNLP community channels or open an issue in the dataset repository.