import json import datasets _AMAZON_REVIEW_2023_DESCRIPTION = """\ Amazon Review 2023 is an updated version of the Amazon Review 2018 dataset. This dataset mainly includes reviews (ratings, text) and item metadata (desc- riptions, category information, price, brand, and images). Compared to the pre- vious versions, the 2023 version features larger size, newer reviews (up to Sep 2023), richer and cleaner meta data, and finer-grained timestamps (from day to milli-second). """ class RawMetaAmazonReview2023Config(datasets.BuilderConfig): def __init__(self, **kwargs): super(RawMetaAmazonReview2023Config, self).__init__(**kwargs) self.suffix = 'jsonl' self.domain = self.name[len(f'raw_meta_'):] self.description = f'This is a subset for items in domain: {self.domain}.' self.data_dir = f'raw/meta_categories/meta_{self.domain}.jsonl' class AmazonReview2023(datasets.GeneratorBasedBuilder): BUILDER_CONFIGS = [ RawMetaAmazonReview2023Config( name='raw_meta_All_Beauty' ) ] def _info(self): return datasets.DatasetInfo( description=_AMAZON_REVIEW_2023_DESCRIPTION + self.config.description ) def _split_generators(self, dl_manager): dl_dir = dl_manager.download_and_extract(self.config.data_dir) return [ datasets.SplitGenerator( name='full', gen_kwargs={"filepath": dl_dir} ) ] def _generate_examples(self, filepath): with open(filepath, 'r', encoding='utf-8') as file: for idx, line in enumerate(file): if self.config.suffix == 'jsonl': try: dp = json.loads(line) """ For item metadata, 'details' is free-form structured data Here we dump it to string to make huggingface datasets easy to store. """ if isinstance(self.config, RawMetaAmazonReview2023Config) and \ 'details' in dp: dp['details'] = json.dumps(dp['details']) except: continue else: raise ValueError(f'Unknown suffix {self.config.suffix}.') yield idx, dp