afriqa.py

"""AfriQA dataset."""


import json
import os
from textwrap import dedent

import datasets


_HOMEPAGE = "https://github.com/masakhane-io/afriqa"

_DESCRIPTION = """\
AfriQA: Cross-lingual Open-Retrieval Question Answering for African Languages

AfriQA is the first cross-lingual question answering (QA) dataset with a focus on African languages. 
The dataset includes over 12,000 XOR QA examples across 10 African languages, making it an invaluable resource for developing more equitable QA technology.
"""

_CITATION = """\

"""

_URL = "https://github.com/masakhane-io/afriqa/raw/main/data/queries/"

_LANG_2_PIVOT = {
    "bem": "en",
    "fon": "fr",
    "hau": "en",
    "ibo": "en",
    "kin": "en",
    "swa": "en",
    "twi": "en",
    "wol": "fr",
    "yor": "en",
    "zul": "en",
}

class AfriQAConfig(datasets.BuilderConfig):
    """BuilderConfig for AfriQA"""

    def __init__(self, **kwargs):
        """BuilderConfig for AfriQA.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(AfriQAConfig, self).__init__(**kwargs)


class AfriQA(datasets.GeneratorBasedBuilder):
    """AfriQA dataset."""

    VERSION = datasets.Version("1.0.0")

    BUILDER_CONFIGS = [
        AfriQAConfig(name="bem", version=datasets.Version("1.0.0"), description="AfriQA Bemba dataset"),
        AfriQAConfig(name="fon", version=datasets.Version("1.0.0"), description="AfriQA Fon dataset"),
        AfriQAConfig(name="hau", version=datasets.Version("1.0.0"), description="AfriQA Hausa dataset"),
        AfriQAConfig(name="ibo", version=datasets.Version("1.0.0"), description="AfriQA Igbo dataset"),
        AfriQAConfig(name="kin", version=datasets.Version("1.0.0"), description="AfriQA Kinyarwanda dataset"),
        AfriQAConfig(name="swa", version=datasets.Version("1.0.0"), description="AfriQA Swahili dataset"),
        AfriQAConfig(name="twi", version=datasets.Version("1.0.0"), description="AfriQA Twi dataset"),
        AfriQAConfig(name="wol", version=datasets.Version("1.0.0"), description="AfriQA Wolof dataset"),
        AfriQAConfig(name="yor", version=datasets.Version("1.0.0"), description="AfriQA Yoruba dataset"),
        AfriQAConfig(name="zul", version=datasets.Version("1.0.0"), description="AfriQA Zulu dataset"),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "question": datasets.Value("string"),
                    "answers": datasets.Value("string"),
                    "lang": datasets.Value("string"),
                    "split": datasets.Value("string"),
                    "translated_question": datasets.Value("string"),
                    "translated_answer": datasets.Value("string"),
                    "translation_type": datasets.Value("string"),
                }
            ),
            homepage=_HOMEPAGE,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        urls_to_download = {
            "train": f"{_URL}{self.config.name}/queries.afriqa.{self.config.name}.{_LANG_2_PIVOT[self.config.name]}.train.json",
            "dev": f"{_URL}{self.config.name}/queries.afriqa.{self.config.name}.{_LANG_2_PIVOT[self.config.name]}.dev.json",
            "test": f"{_URL}{self.config.name}/queries.afriqa.{self.config.name}.{_LANG_2_PIVOT[self.config.name]}.test.json",
        }
        downloaded_files = dl_manager.download_and_extract(urls_to_download)

        return [
            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}),
            datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["dev"]}),
            datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test"]}),
        ]

    def _generate_examples(self, filepath):
        """Yields examples."""
        with open(filepath, encoding="utf-8-sig") as f:
            for _, row in enumerate(f):
                example = json.loads(row)
                _id = example["id"]

                yield _id, {
                    "question": example["question"],
                    "answers": example["answers"],
                    "lang": example["lang"],
                    "split": example["split"],
                    "translated_question": example["translated_question"],
                    "translated_answer": example["translated_answer"],
                    "translation_type": example["translation_type"],
                }