Update EpiSet4NER-v2.py

EpiSet4NER-v2.py

@@ -1,5 +1,5 @@
 # coding=utf-8
-# Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
+# Copyright 2020 HuggingFace Datasets Authors.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -14,61 +14,122 @@
 # limitations under the License.
 
 # Lint as: python3
-"""EpiClassify4GARD dataset."""
+"""INSERT TITLE"""
 
+import logging
 
-import csv
 import datasets
-from datasets.tasks import TextClassification
 
 
-_DESCRIPTION = """\
-INSERT DESCRIPTION
-"""
 _CITATION = """\
-John JN, Sid E, Zhu Q. Recurrent Neural Networks to Automatically Identify Rare Disease Epidemiologic Studies from PubMed. AMIA Jt Summits Transl Sci Proc. 2021 May 17;2021:325-334. PMID: 34457147; PMCID: PMC8378621.
+*REDO*
+@inproceedings{wang2019crossweigh,
+ title={CrossWeigh: Training Named Entity Tagger from Imperfect Annotations},
+ author={Wang, Zihan and Shang, Jingbo and Liu, Liyuan and Lu, Lihao and Liu, Jiacheng and Han, Jiawei},
+ booktitle={Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)},
+ pages={5157--5166},
+ year={2019}
+}
+"""
+
+_DESCRIPTION = """\
+**REWRITE*
+EpiSet4NER is a dataset generated from 620 rare disease abstracts labeled using statistical and rule-base methods. The test set was then manually corrected by a rare disease expert.
+For more details see *INSERT PAPER* and https://github.com/ncats/epi4GARD/tree/master/EpiExtract4GARD#epiextract4gard
 """
 
-_TRAIN_DOWNLOAD_URL = "https://huggingface.co/datasets/ncats/GARD_EpiSet4TextClassification/raw/main/train_short.tsv"
-_VAL_DOWNLOAD_URL = "https://huggingface.co/datasets/ncats/GARD_EpiSet4TextClassification/raw/main/val_short.tsv"
-_TEST_DOWNLOAD_URL = "https://huggingface.co/datasets/ncats/GARD_EpiSet4TextClassification/raw/main/test.tsv"
+_URL = "https://github.com/NCATS/epi4GARD/raw/master/EpiExtract4GARD/datasets/EpiCustomV3/"
+_TRAINING_FILE = "train.tsv"
+_VAL_FILE = "val.tsv"
+_TEST_FILE = "test.tsv"
 
 
-class EpiClassify4GARD(datasets.GeneratorBasedBuilder):
- """EpiClassify4GARD text classification dataset."""
+class EpiSetConfig(datasets.BuilderConfig):
+ """BuilderConfig for Conll2003"""
+
+ def __init__(self, **kwargs):
+ """BuilderConfig forConll2003.
+ Args:
+ **kwargs: keyword arguments forwarded to super.
+ """
+ super(EpiSetConfig, self).__init__(**kwargs)
+
+
+class EpiSet(datasets.GeneratorBasedBuilder):
+ """EpiSet4NER by GARD."""
+
+ BUILDER_CONFIGS = [
+ EpiSetConfig(name="EpiSet4NER", version=datasets.Version("3.2.1"), description="EpiSet4NER by NIH NCATS GARD"),
+ ]
 
  def _info(self):
  return datasets.DatasetInfo(
  description=_DESCRIPTION,
  features=datasets.Features(
  {
- "abstract": datasets.Value("string"),
- "label": datasets.features.ClassLabel(names=["1 = IsEpi", "0 = IsNotEpi"]),
+ "id": datasets.Value("string"),
+ "tokens": datasets.Sequence(datasets.Value("string")),
+ "ner_tags": datasets.Sequence(
+ datasets.features.ClassLabel(
+ names=[
+ "O", #(0)
+ "B-LOC", #(1)
+ "I-LOC", #(2)
+ "B-EPI", #(3)
+ "I-EPI", #(4)
+ "B-STAT", #(5)
+ "I-STAT", #(6)
+ ]
+ )
+ ),
  }
  ),
- homepage="https://github.com/ncats/epi4GARD/tree/master/Epi4GARD#epi4gard",
+ supervised_keys=None,
+ homepage="https://github.com/ncats/epi4GARD/tree/master/EpiExtract4GARD#epiextract4gard",
  citation=_CITATION,
- task_templates=[TextClassification(text_column="abstract", label_column="label")],
  )
 
  def _split_generators(self, dl_manager):
- train_path = dl_manager.download_and_extract(_TRAIN_DOWNLOAD_URL)
- val_path = dl_manager.download_and_extract(_VAL_DOWNLOAD_URL)
- test_path = dl_manager.download_and_extract(_TEST_DOWNLOAD_URL)
+ """Returns SplitGenerators."""
+ urls_to_download = {
+ "train": f"{_URL}{_TRAINING_FILE}",
+ "val": f"{_URL}{_VAL_FILE}",
+ "test": f"{_URL}{_TEST_FILE}",
+ }
+ downloaded_files = dl_manager.download_and_extract(urls_to_download)
+
  return [
- datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": train_path}),
- datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": val_path }),
- datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": test_path}),
+ datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}),
+ datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["val"]}),
+ datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test"]}),
  ]
 
  def _generate_examples(self, filepath):
- """Generate examples."""
- with open(filepath, encoding="utf-8") as csv_file:
- csv_reader = csv.reader(
- csv_file, quotechar='"', delimiter="\t", quoting=csv.QUOTE_ALL, skipinitialspace=True
- )
- next(csv_reader)
- for id_, row in enumerate(csv_reader):
- abstract = row[0]
- label = row[1]
- yield id_, {"abstract": abstract, "label": int(label)}
+ logging.info("⏳ Generating examples from = %s", filepath)
+ with open(filepath, encoding="utf-8") as f:
+ guid = 0
+ tokens = []
+ ner_tags = []
+ for line in f:
+ if line.startswith("-DOCSTART-") or line == "" or line == "\n":
+ if tokens:
+ yield guid, {
+ "id": str(guid),
+ "tokens": tokens,
+ "ner_tags": ner_tags,
+ }
+ guid += 1
+ tokens = []
+ ner_tags = []
+ else:
+ # EpiSet tokens are space separated
+ splits = line.split("\t")
+ tokens.append(splits[0])
+ ner_tags.append(splits[1].rstrip())
+ # last example
+ if tokens:
+ yield guid, {
+ "id": str(guid),
+ "tokens": tokens,
+ "ner_tags": ner_tags,
+ }