Estwld commited on
Commit
c1b0800
1 Parent(s): 5c07077

Upload 2 files

Browse files
Files changed (2) hide show
  1. constants.py +37 -0
  2. make_origin_dataset.py +116 -0
constants.py ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ relations_map = {
2
+ 'social_intercation': ['oEffect', 'oReact', 'oWant', 'xAttr', 'xEffect', 'xIntent', 'xNeed', 'xReact', 'xWant'],
3
+ 'physical_entity' : ['ObjectUse', 'AtLocation', 'Desires', 'NotDesires', 'HasProperty', 'MadeUpOf', 'CapableOf'],
4
+ 'event_centered' : ['isAfter', 'isBefore', 'HasSubEvent', 'HinderedBy', 'Causes', 'xReason', 'isFilledBy'],
5
+ }
6
+
7
+ social_intercation_2_descriptions = {
8
+ 'xIntent': 'Why does X cause the event?',
9
+ 'xNeed' : 'What does X need to do before the event?',
10
+ 'xAttr' : 'How would X be described?',
11
+ 'xEffect': 'What effects does the event have on X?',
12
+ 'xWant' : 'What would X likely want to do after the event?',
13
+ 'xReact' : 'How does X feel after the event?',
14
+ 'oEffect': 'What effects does the event have on others?',
15
+ 'oReact' : 'How do others\' feel after the event?',
16
+ 'oWant' : 'What would others likely want to do after the event?',
17
+ }
18
+
19
+ physical_entity_2_descriptions = {
20
+ 'ObjectUse' : 'ObjectUse describes everyday affordances or uses of objects, and includes both typical and atypical uses.',
21
+ 'AtLocation' : 'AtLocation is a spatial relation that describes the location in/on/at which an entity is likely to be found.',
22
+ 'Desires' : 'Desires are relations that deal with desires of sentient entities.',
23
+ 'HasProperty': 'HasProperty usually describes entities\' general characteristics. In certain case, the relation can also map to descriptors that speak to the substance or value of items.',
24
+ 'NotDesires' : 'Not desires are relations that deal with the absence of desires in sentient entities.',
25
+ 'MadeUpOf' : 'MadeUpOf describes a part, portion or makeup of an entity.',
26
+ 'CapableOf' : 'CapableOf describes abilities and capabilities of everyday living entities and natural entites that can exert a force. CapableOf includes general capabilities and specialized capabilities.',
27
+ }
28
+
29
+ event_centered_2_descriptions = {
30
+ 'isAfter' : 'isAfter introduces events that can follow an event.',
31
+ 'HasSubEvent': 'HasSubEvent provides the internal structure of an event, each tail denoting a step within the larger head event.',
32
+ 'isBefore' : 'isBefore introduces events that can precede an event.',
33
+ 'HinderedBy' : 'HinderedyBy introduces hindrances that obstruct the natural path to the achievement of a goal.',
34
+ 'Causes' : 'Causes specifically captures the causal relation between two events or entities.',
35
+ 'xReason' : 'xReason provides a post-fact explanation of the cause of an event.',
36
+ 'isFilledBy' : 'isFilledBy provides a filler phrase for an event with a blank that is sensical and commonly acceptable for the event.',
37
+ }
make_origin_dataset.py ADDED
@@ -0,0 +1,116 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from collections import Counter
2
+ from typing import List
3
+
4
+ import datasets
5
+ import matplotlib.pyplot as plt
6
+ import pandas as pd
7
+ from constants import (event_centered_2_descriptions,
8
+ physical_entity_2_descriptions, relations_map,
9
+ social_intercation_2_descriptions)
10
+
11
+
12
+ def show_bar(relation: List):
13
+ c = dict(Counter(relation))
14
+
15
+ keys = list(c.keys())
16
+ values = list(c.values())
17
+ plt.bar(keys, values)
18
+ plt.xticks(rotation=25, fontsize=8)
19
+ plt.yticks(fontsize=8)
20
+
21
+ plt.xlabel('relations')
22
+ plt.ylabel('numbers')
23
+ plt.title('relations analysis')
24
+
25
+ for i in range(len(keys)):
26
+ plt.text(i, values[i] + 10, str(values[i]), ha='center', fontsize=10)
27
+
28
+ plt.show()
29
+
30
+ def read_file(data_path: str):
31
+ df = pd.read_csv(data_path, sep='\t', header=None)
32
+
33
+ df.columns = ['event', 'relation', 'tail']
34
+ print(df.head())
35
+
36
+ event = df['event'].tolist()
37
+ relation = df['relation'].tolist()
38
+ tail = df['tail'].tolist()
39
+
40
+ return event, relation, tail
41
+
42
+ def make_base_dataset(event: List[str], relation: List[str], tail: List[str]):
43
+ new_event, new_relation, new_tail = [], [], []
44
+ knowledge_type = []
45
+ relation_description = []
46
+
47
+ prev_event, prev_relation, prev_tail = None, None, None
48
+
49
+ for i in range(len(event)):
50
+ if i > 0 and event[i] == prev_event and relation[i] == prev_relation:
51
+ new_tail[-1].extend( [tail[i]] )
52
+ else:
53
+ new_event.append(event[i])
54
+ new_relation.append(relation[i])
55
+ # insert knowledge type
56
+ relation_list = []
57
+ for r in list(relations_map.values()):
58
+ relation_list.extend(r)
59
+ if relation[i] not in relation_list:
60
+ raise ValueError(f'dont find match knowledge type named {relation[i]}, please check it!')
61
+
62
+ for k, v in relations_map.items():
63
+ if relation[i] in v:
64
+ knowledge_type.append(k)
65
+ if k == 'social_intercation':
66
+ relation_description.append(social_intercation_2_descriptions[relation[i]])
67
+ elif k == 'physical_entity':
68
+ relation_description.append(physical_entity_2_descriptions[relation[i]])
69
+ elif k == 'event_centered':
70
+ relation_description.append(event_centered_2_descriptions[relation[i]])
71
+ else:
72
+ raise KeyError(f"dont find match relation type named {relation[i]} in dict, please check it!")
73
+
74
+ new_tail.append( [tail[i]] )
75
+ prev_event, prev_relation, prev_tail = event[i], relation[i], tail[i]
76
+
77
+ df = pd.DataFrame({
78
+ 'knowledge_type': knowledge_type,
79
+ 'event': new_event,
80
+ 'relation': new_relation,
81
+ 'relation_description': relation_description,
82
+ 'tail': new_tail,
83
+ })
84
+
85
+ print(df.head())
86
+ return df
87
+
88
+ def get_dataset(data_path: str):
89
+ event, relation, tail = read_file(data_path=data_path)
90
+ df = make_base_dataset(event=event, relation=relation, tail=tail)
91
+ dataset = datasets.Dataset.from_pandas(df, split='train')
92
+
93
+ print(dataset)
94
+ return dataset
95
+
96
+ def upload_dataset(dataset, repo_id :str, access_token: str, private: bool):
97
+ dataset.push_to_hub(
98
+ repo_id = repo_id,
99
+ private = private,
100
+ token = access_token,
101
+ )
102
+
103
+ if __name__ == '__main__':
104
+ train_dataset = get_dataset('./dataset/train.tsv')
105
+ valid_dataset = get_dataset('./dataset/dev.tsv')
106
+ test_dataset = get_dataset('./dataset/test.tsv')
107
+
108
+ dataset = datasets.DatasetDict({
109
+ 'train': train_dataset,
110
+ 'validation': valid_dataset,
111
+ 'test': test_dataset
112
+ })
113
+
114
+ print(dataset)
115
+
116
+ upload_dataset(dataset, repo_id='Estwld/atomic2020-origin', private=False, access_token='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')