liyucheng commited on
Commit
1edb94c
1 Parent(s): 1c7bce1

Upload chinese_metaphor_dataset.py

Browse files
Files changed (1) hide show
  1. chinese_metaphor_dataset.py +85 -0
chinese_metaphor_dataset.py ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import datasets
2
+ # import pandas as pd
3
+ import json
4
+
5
+ _CITATION = """\
6
+ @inproceedings{li-etal-2022-cm,
7
+ title = "{CM}-Gen: A Neural Framework for {C}hinese Metaphor Generation with Explicit Context Modelling",
8
+ author = "Li, Yucheng and
9
+ Lin, Chenghua and
10
+ Guerin, Frank",
11
+ booktitle = "Proceedings of the 29th International Conference on Computational Linguistics",
12
+ month = oct,
13
+ year = "2022",
14
+ address = "Gyeongju, Republic of Korea",
15
+ publisher = "International Committee on Computational Linguistics",
16
+ url = "https://aclanthology.org/2022.coling-1.563",
17
+ pages = "6468--6479",
18
+ }
19
+
20
+ @misc{li-inlg-2022-nominal,
21
+ doi = {10.48550/ARXIV.2206.05195},
22
+ url = {https://arxiv.org/abs/2206.05195},
23
+ author = {Li, Yucheng and Lin, Chenghua and Geurin, Frank},
24
+ keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
25
+ title = {Nominal Metaphor Generation with Multitask Learning},
26
+ publisher = {arXiv},
27
+ year = {2022},
28
+ copyright = {arXiv.org perpetual, non-exclusive license}
29
+ }
30
+ """
31
+
32
+ _DESCRIPTION = """\
33
+ Chinese Metaphor Corpus
34
+
35
+ The first Chinese metaphor corpus serving both metaphor identification and generation.
36
+ 首个中文比喻数据集,可以用于中文比喻识别与中文比喻生成。
37
+ """
38
+
39
+ _HOMEPAGE = "https://github.com/liyucheng09/Metaphor_Generator"
40
+
41
+ _URL = 'https://github.com/liyucheng09/Metaphor_Generator/raw/master/CMC/zh_CMC2.txt'
42
+
43
+ class CMC(datasets.GeneratorBasedBuilder):
44
+
45
+ BUILDER_CONFIG_CLASS=datasets.BuilderConfig
46
+
47
+ def _info(self):
48
+
49
+ feature = {
50
+ 'sent': datasets.Value('string'),
51
+ 'tenor': datasets.Value('string'),
52
+ 'comparator': datasets.Value('string'),
53
+ 'vehicle':datasets.Value('string'),
54
+ }
55
+
56
+ return datasets.DatasetInfo(
57
+ description=_DESCRIPTION,
58
+ features=datasets.Features(feature),
59
+ homepage=_HOMEPAGE,
60
+ citation=_CITATION
61
+ )
62
+
63
+ def _split_generators(self, dl_manager):
64
+
65
+ cmc_file = dl_manager.download(_URL)
66
+
67
+ return datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={'cmc_file': cmc_file}),
68
+ # return [
69
+ # datasets.SplitGenerator(name='hard', gen_kwargs={'filepath': self.config.data_path}),
70
+ # ]
71
+
72
+ def _generate_examples(self, cmc_file):
73
+ with open(cmc_file, encoding='utf-8') as f:
74
+ count = 0
75
+ for line in f:
76
+ if not line:
77
+ continue
78
+ data = json.loads(line)
79
+ yield count, {
80
+ 'sent': data['sent'],
81
+ 'tenor': data['tenor'],
82
+ 'comparator': data['comparator'],
83
+ 'vehicle': data['vehicle']
84
+ }
85
+ count+=1