BESSTIE.py

import os

import datasets
from typing import List
import json

logger = datasets.logging.get_logger(__name__)


_CITATION = """
"""

_DESCRIPTION = """
This is the dataset repository for BESSTIE Dataset.
The dataset can help build text classification models for sarcasm detection and sentiment analysis for low resource languages.
"""

class BESSTIEConfig(datasets.BuilderConfig):
    """BuilderConfig for BESSTIE"""

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


class BESSTIEConfig(datasets.GeneratorBasedBuilder):
    """BESSTIE dataset."""

    BUILDER_CONFIGS = [
        BESSTIEConfig(name="BESSTIE", version=datasets.Version("0.0.1"), description="BESSTIE dataset"),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "id": datasets.Value("int32"),
                    "text": datasets.Value("string"),
                    "sentiment_label": datasets.Value("int32"),
                }
            ),
            supervised_keys=None,
            citation=_CITATION,
        )

    _URL = "https://huggingface.co./datasets/mindhunter23/BESSTIE-google-sentiment-au/blob/main/data/"
    _URLS = {
        "train": _URL + "google-sentiment-au-train.jsonl",
        "dev": _URL + "google-sentiment-au-valid.jsonl",
        # "test": _URL + "google-sentiment-au-test.jsonl"
    }

    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
        urls_to_download = self._URLS
        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):
        """This function returns the examples in the raw (text) form."""
        logger.info("generating examples from = %s", filepath)
        with open(filepath) as f:
            for line in f:
                object = json.loads(line.strip())
                id_ = int(object['id'])
                yield id_, {
                    "id": id_,
                    "text": object['text'],
                    "sentiment_label": int(object['sentiment_label']),
                }