File size: 4,380 Bytes
d690744 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
import re
import os
import logging
import pickle
import argparse
import numpy as np
import pandas as pd
from tqdm import tqdm
def make_summarization_csv(args):
if args.for_qfid:
logging.info('Making csv files for QFiD...')
logging.info('Columns={"reference": literature review title <s> chapter title </s> literature review title <s> chapter title <s> abstract of cited paper 1 <s> BIB001 </s> literature review title <s> chapter title <s> abstract of cited paper 2 <s> BIB002 </s> ..., "target": literature review chapter}')
else:
logging.info('Making csv files for summarization...')
logging.info('Columns={"reference": literature review title <s> chapter title <s> abstract of cited paper 1 <s> BIB001 </s> literature review title <s> chapter title <s> abstract of cited paper 2 <s> BIB002 </s> ..., "target": literature review chapter}')
section_df = pd.read_pickle(os.path.join(args.dataset_path, 'split_survey_df.pkl'))
dataset_df = section_df[section_df['n_bibs'].apply(lambda n_bibs: n_bibs >= 2)]
dataset_df = dataset_df.rename(columns={'text': 'target'})
dataset_df = dataset_df.rename(columns={'bib_cinting_sentences': 'bib_citing_sentences'})
dataset_df['reference'] = dataset_df[['bib_abstracts', 'section', 'title']].apply(lambda bib_abstracts: ' '.join(['</s> {} <s> {} <s> {} <s> BIB{}'.format(bib_abstracts[2], bib_abstracts[1], abstract, bib) for bib, abstract in bib_abstracts[0].items()]), axis=1)
if args.for_qfid:
dataset_df['reference'] = dataset_df['title'] + ' <s> ' + dataset_df['section'] + ' ' + dataset_df['reference']
else:
dataset_df['reference'] = dataset_df['reference'].apply(lambda s: s[5:])
split_df = dataset_df['split']
dataset_df = dataset_df[['reference', 'target']]
train_df = dataset_df[split_df == 'train']
val_df = dataset_df[split_df == 'val']
test_df = dataset_df[split_df == 'test']
if args.for_qfid:
train_df.to_csv(os.path.join(args.dataset_path, 'train_qfid.csv'), index=False)
val_df.to_csv(os.path.join(args.dataset_path, 'val_qfid.csv'), index=False)
test_df.to_csv(os.path.join(args.dataset_path, 'test_qfid.csv'), index=False)
else:
train_df.to_csv(os.path.join(args.dataset_path, 'train.csv'), index=False)
val_df.to_csv(os.path.join(args.dataset_path, 'val.csv'), index=False)
test_df.to_csv(os.path.join(args.dataset_path, 'test.csv'), index=False)
logging.info('Done!')
def anonymize_bib(args):
logging.info('Converting BIB identifiers...')
for split in ['val', 'test', 'train']:
if args.for_qfid:
df = pd.read_csv(os.path.join(args.dataset_path, '{}_qfid.csv'.format(split)))
else:
df = pd.read_csv(os.path.join(args.dataset_path, '{}.csv'.format(split)))
bar = tqdm(total=len(df))
for row in df.itertuples():
cnt = 1
bib_dict = {}
for i in range(len(row.reference)):
if row.reference[i:i+7] == '<s> BIB':
bib_dict[row.reference[i+7:].split(' ')[0]] = cnt
cnt += 1
ref = row.reference
tgt = row.target
for key, value in bib_dict.items():
ref = re.sub('BIB{}'.format(key), 'BIB{:0>3}'.format(value), ref)
tgt = re.sub('BIB{}'.format(key), 'BIB{:0>3}'.format(value), tgt)
df.at[row.Index, 'reference'] = ref
df.at[row.Index, 'target'] = tgt
bar.update(1)
logging.info('Saving...')
if args.for_qfid:
df.to_csv(os.path.join(args.dataset_path, '{}_qfid.csv'.format(split)), index=False)
else:
df.to_csv(os.path.join(args.dataset_path, '{}.csv'.format(split)), index=False)
if __name__ == '__main__':
logging.basicConfig(format='%(message)s', level=logging.DEBUG)
parser = argparse.ArgumentParser(description='')
parser.add_argument('-dataset_path', default=".", help='Path to the generated dataset')
parser.add_argument('--for_qfid', action='store_true', help='Add if you train QFiD on the generated csv files')
args = parser.parse_args()
make_summarization_csv(args) # Convert split_survey_df into csv files suitable for summarization
anonymize_bib(args) # Converting BIB{paper_id} into BIB{001, 002, ...}
|