wrime-sentiment.py

import csv
from dataclasses import dataclass
from typing import Any, Dict, List, Tuple

import datasets

logger = datasets.logging.get_logger(__name__)

_DESCRIPTION = (
    "日本語の感情分析データセット WRIME を、ポジティブ/ネガティブの二値分類のタスクに加工したデータセットです。"
    "GitHub リポジトリ ids-cv/wrime で公開されているデータセットを利用しています。"
)

_URL = "https://raw.githubusercontent.com/ids-cv/wrime/master/wrime-ver2.tsv"


@dataclass
class WrimeSentimentConfig(datasets.BuilderConfig):
    remove_neutral: bool = True


class WrimeSentiment(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIG_CLASS = WrimeSentimentConfig

    def _info(self) -> datasets.DatasetInfo:
        labels = ["positive", "negative"]
        if not self.config.remove_neutral:
            labels += ["neutral"]
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "sentence": datasets.Value("string"),
                    "label": datasets.ClassLabel(
                        num_classes=len(labels), names=labels
                    ),
                    "user_id": datasets.Value("int64"),
                    "datetime": datasets.Value("string")
                }
            ),
        )

    def _split_generators(
        self, dl_manager: datasets.DownloadManager
    ) -> List[datasets.SplitGenerator]:
        downloaded_file = dl_manager.download_and_extract(_URL)
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "filepath": downloaded_file,
                    "split": "train",
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                gen_kwargs={
                    "filepath": downloaded_file,
                    "split": "dev",
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "filepath": downloaded_file,
                    "split": "test",
                },
            ),
        ]

    def _generate_examples(
        self, filepath: str, split: str
    ) -> Tuple[str, Dict[str, Any]]:
        logger.info(f"generating examples from {filepath}")
        _key = 0

        with open(filepath, "r", encoding="utf-8") as f:
            reader = csv.DictReader(f, delimiter="\t")
            for data in reader:
                if data["Train/Dev/Test"].lower() != split:
                    continue

                sentiment_score = int(data["Avg. Readers_Sentiment"])
                if sentiment_score > 0:
                    label = "positive"
                elif sentiment_score < 0:
                    label = "negative"
                else:
                    label = "neutral"

                if self.config.remove_neutral and label == "neutral":
                    continue
                yield _key, {
                    "sentence": data["Sentence"],
                    "label": label,
                    "user_id": data["UserID"],
                    "datetime": data["Datetime"].strip(),
                }
                _key += 1