from pathlib import Path from typing import List import datasets from seacrowd.utils import schemas from seacrowd.utils.configs import SEACrowdConfig from seacrowd.utils.constants import Tasks _CITATION = """\ @inproceedings{van-der-goot-etal-2020-cross, title={From Masked-Language Modeling to Translation: Non-{E}nglish Auxiliary Tasks Improve Zero-shot Spoken Language Understanding}, author={van der Goot, Rob and Sharaf, Ibrahim and Imankulova, Aizhan and {\"U}st{\"u}n, Ahmet and Stepanovic, Marija and Ramponi, Alan and Khairunnisa, Siti Oryza and Komachi, Mamoru and Plank, Barbara}, booktitle = "Proceedings of the 2021 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)", year = "2021", address = "Mexico City, Mexico", publisher = "Association for Computational Linguistics" } """ _DATASETNAME = "xsid" _DESCRIPTION = """\ XSID is a new benchmark for cross-lingual (X) Slot and Intent Detection in 13 languages from 6 language families, including a very low-resource dialect. """ _HOMEPAGE = "https://bitbucket.org/robvanderg/xsid/src/master/" _LANGUAGES = ["ind"] _LICENSE = "CC-BY-SA 4.0" _LOCAL = False _URLS = { _DATASETNAME: "https://bitbucket.org/robvanderg/xsid/get/04ce1e6c8c28.zip", } _SUPPORTED_TASKS = [Tasks.INTENT_CLASSIFICATION, Tasks.POS_TAGGING] _SOURCE_VERSION = "0.3.0" _SEACROWD_VERSION = "2024.06.20" INTENT_LIST = [ "AddToPlaylist", "BookRestaurant", "PlayMusic", "RateBook", "SearchCreativeWork", "SearchScreeningEvent", "alarm/cancel_alarm", "alarm/modify_alarm", "alarm/set_alarm", "alarm/show_alarms", "alarm/snooze_alarm", "alarm/time_left_on_alarm", "reminder/cancel_reminder", "reminder/set_reminder", "reminder/show_reminders", "weather/checkSunrise", "weather/checkSunset", "weather/find" ] TAG_LIST = [ "B-album", "B-artist", "B-best_rating", "B-condition_description", "B-condition_temperature", "B-cuisine", "B-datetime", "B-ecurring_datetime", "B-entity_name", "B-facility", "B-genre", "B-location", "B-movie_name", "B-movie_type", "B-music_item", "B-object_location_type", "B-object_name", "B-object_part_of_series_type", "B-object_select", "B-object_type", "B-party_size_description", "B-party_size_number", "B-playlist", "B-rating_unit", "B-rating_value", "B-recurring_datetime", "B-reference", "B-reminder/todo", "B-restaurant_name", "B-restaurant_type", "B-served_dish", "B-service", "B-sort", "B-track", "B-weather/attribute", "I-album", "I-artist", "I-best_rating", "I-condition_description", "I-condition_temperature", "I-cuisine", "I-datetime", "I-ecurring_datetime", "I-entity_name", "I-facility", "I-genre", "I-location", "I-movie_name", "I-movie_type", "I-music_item", "I-object_location_type", "I-object_name", "I-object_part_of_series_type", "I-object_select", "I-object_type", "I-party_size_description", "I-party_size_number", "I-playlist", "I-rating_unit", "I-rating_value", "I-recurring_datetime", "I-reference", "I-reminder/todo", "I-restaurant_name", "I-restaurant_type", "I-served_dish", "I-service", "I-sort", "I-track", "I-weather/attribute", "O", "Orecurring_datetime" ] class XSID(datasets.GeneratorBasedBuilder): """xSID datasets contains datasets to detect the intent from the text""" BUILDER_CONFIGS = [ SEACrowdConfig( name="xsid_source", version=datasets.Version(_SOURCE_VERSION), description="xSID source schema", schema="source", subset_id="xsid", ), SEACrowdConfig( name="xsid_seacrowd_text", version=datasets.Version(_SEACROWD_VERSION), description="xSID Nusantara intent classification schema", schema="seacrowd_text", subset_id="xsid", ), SEACrowdConfig( name="xsid_seacrowd_seq_label", version=datasets.Version(_SEACROWD_VERSION), description="xSID Nusantara pos tagging schema", schema="seacrowd_seq_label", subset_id="xsid", ), ] DEFAULT_CONFIG_NAME = "xsid_source" def _info(self) -> datasets.DatasetInfo: if self.config.schema == "source": features = datasets.Features( { "id": datasets.Value("string"), "text": datasets.Value("string"), "text-en": datasets.Value("string"), "intent": datasets.Value("string"), "tokens": datasets.Sequence(datasets.Value("string")), } ) elif self.config.schema == "seacrowd_text": features = schemas.text_features(label_names=INTENT_LIST) elif self.config.schema == "seacrowd_seq_label": features = schemas.seq_label_features(label_names=TAG_LIST) else: raise ValueError(f"Invalid config schema: {self.config.schema}") return datasets.DatasetInfo( description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION, ) def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]: urls = _URLS[_DATASETNAME] base_path = Path(dl_manager.download_and_extract(urls)) / "robvanderg-xsid-04ce1e6c8c28" / "data" / "xSID-0.3" data_files = { "train": base_path / "id.projectedTrain.conll", "test": base_path / "id.test.conll", "validation": base_path / "id.valid.conll" } return [ datasets.SplitGenerator( name=datasets.Split.TRAIN, gen_kwargs={"filepath": data_files["train"]}, ), datasets.SplitGenerator( name=datasets.Split.TEST, gen_kwargs={"filepath": data_files["test"]}, ), datasets.SplitGenerator( name=datasets.Split.VALIDATION, gen_kwargs={"filepath": data_files["validation"]}, ), ] def _generate_examples(self, filepath: Path): print('filepath', filepath) if self.config.name == "xsid_source": with open(filepath, "r") as file: data = file.read().strip("\n").split("\n\n") i = 0 for sample in data: id = "" tokens = [] for row_sample in sample.split("\n"): s = row_sample.split(": ") if s[0] == "# id": id = s[1] elif s[0] == "# text-en": text_en = s[1] elif s[0] == "# text": text = s[1] elif s[0] == "# intent": intent = s[1] else: tokens.append(s[0]) if id == "": id = i i = i + 1 ex = { "id": id, "text": text, "text-en": text_en, "intent": intent, "tokens": tokens } yield id, ex elif self.config.name == "xsid_seacrowd_text": with open(filepath, "r") as file: data = file.read().strip("\n").split("\n\n") i = 0 for sample in data: id = "" for row_sample in sample.split("\n"): s = row_sample.split(": ") if s[0] == "# id": id = s[1] elif s[0] == "# text": text = s[1] elif s[0] == "# intent": intent = s[1] if id == "": id = i i = i + 1 ex = { "id": id, "text": text, "label": intent } yield id, ex elif self.config.name == "xsid_seacrowd_seq_label": with open(filepath, "r") as file: data = file.read().strip("\n").split("\n\n") i = 0 for sample in data: id = "" tokens = [] labels = [] for row_sample in sample.split("\n"): s = row_sample.split(": ") if s[0] == "# id": id = s[1] elif len(s) == 1: tokens.append(s[0].split("\t")[1]) labels.append(s[0].split("\t")[3]) if id == "": id = i i = i + 1 ex = { "id": id, "tokens": tokens, "labels": labels } yield id, ex else: raise ValueError(f"Invalid config: {self.config.name}")