mbib-base.py

import datasets
import pandas as pd


_MBIB_DESCRIPTION = 'bla bla'

class MBIBConfig(datasets.BuilderConfig):
    def __init__(self,data_dir,**kwargs):
        super(MBIBConfig, self).__init__(version=datasets.Version("1.0.0", ""), **kwargs)
        self.data_dir = data_dir


class MBIB(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        MBIBConfig(name="cognitive-bias",data_dir="mbib-aggregated/cognitive-bias.csv"),
        MBIBConfig(name="fake-news",data_dir="mbib-aggregated/fake-news.csv"),
        MBIBConfig(name="gender-bias",data_dir="mbib-aggregated/gender-bias.csv"),
        MBIBConfig(name="hate-speech",data_dir="mbib-aggregated/hate-speech.csv"),
        MBIBConfig(name="linguistic-bias",data_dir="mbib-aggregated/linguistic-bias.csv"),
        MBIBConfig(name="political-bias",data_dir="mbib-aggregated/political-bias.csv"),
        MBIBConfig(name="racial-bias",data_dir="mbib-aggregated/racial-bias.csv"),
        MBIBConfig(name="text-level-bias",data_dir="mbib-aggregated/text-level-bias.csv")]
    
    def _info(self):
        features=datasets.Features(
            {"text":datasets.Value("string"),
            "label":datasets.Value("int32")}
        )

        return datasets.DatasetInfo(
            description=_MBIB_DESCRIPTION,
            features=features
        )

    def _split_generators(self, dl_manager):
        data_file = dl_manager.download(self.config.data_dir)
        return [datasets.SplitGenerator(name=datasets.Split.TRAIN,gen_kwargs={"data_dir": data_file})]


    def _generate_examples(self, data_dir):
       df = pd.read_csv(data_dir)[['text','label']]
        
       for i, row in df.iterrows():
                yield i, {"text":row['text'],"label":row['label']}