-
Notifications
You must be signed in to change notification settings - Fork 1
/
get_relations.py
150 lines (101 loc) · 5.39 KB
/
get_relations.py
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
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
import pandas as pd
import glob
import json
import collections
def recursive_dict():
return collections.defaultdict(recursive_dict)
def get_relations_after_2016():
relations_dict = recursive_dict()
names = ["id", "form", "lemma", "upostag", "xpostag", "feats", "head", "deprel", "deps", "misc"]
buzzwords = pd.read_table("buzzword/aged.lexicon/target.test.txt", names=["token", "class"])
pd.options.mode.chained_assignment = None
for filename in glob.glob("stanford_output/*_raw_after_2016.txt*.conllu"):
print(filename)
data = pd.read_table(filename, names=names)
idx = data.index[(data['id'] == 1) & (~data['form'].isin(['！', '？', '。', '.']))]
for i in range(0, len(idx) - 1):
current_post = data[idx[i]:idx[i+1]]
head_dict = pd.Series(current_post['form'].values, index=current_post['id']).to_dict()
for index, row in current_post.iterrows():
if row['head'] != 0:
try:
head_id = int(row['head'])
current_post.loc[index, 'head'] = head_dict[head_id]
except:
pass
for index, row in current_post.iterrows():
if row['head'] in buzzwords['token'].values:
buzz = row['head']
form = row['form']
deprel = row['deprel']
if relations_dict[buzz]["token_relations"]:
relations_dict[buzz]["token_relations"].append(form)
else:
relations_dict[buzz]["token_relations"] = [form]
if relations_dict[buzz]["syntax_relations"]:
relations_dict[buzz]["syntax_relations"].append(deprel)
else:
relations_dict[buzz]["syntax_relations"] = [deprel]
with open("after_2016_relations_raw.json", 'w', encoding='utf8') as json_file:
json.dump(relations_dict, json_file, ensure_ascii=False)
return relations_dict
def get_relations_before_2016():
relations_dict = recursive_dict()
names = ["id", "form", "lemma", "upostag", "xpostag", "feats", "head", "deprel", "deps", "misc"]
buzzwords = pd.read_table("buzzword/aged.lexicon/target.test.txt", names=["token", "class"])
pd.options.mode.chained_assignment = None
for filename in glob.glob("stanford_output/*_raw_before_2016.txt*.conllu"):
print(filename)
data = pd.read_table(filename, names=names)
idx = data.index[(data['id'] == 1) & (~data['form'].isin(['！', '？', '。', '.']))]
for i in range(0, len(idx) - 1):
current_post = data[idx[i]:idx[i+1]]
head_dict = pd.Series(current_post['form'].values, index=current_post['id']).to_dict()
for index, row in current_post.iterrows():
if row['head'] != 0:
try:
head_id = int(row['head'])
current_post.loc[index, 'head'] = head_dict[head_id]
except:
pass
for index, row in current_post.iterrows():
if row['head'] in buzzwords['token'].values:
buzz = row['head']
form = row['form']
deprel = row['deprel']
if relations_dict[buzz]["token_relations"]:
relations_dict[buzz]["token_relations"].append(form)
else:
relations_dict[buzz]["token_relations"] = [form]
if relations_dict[buzz]["syntax_relations"]:
relations_dict[buzz]["syntax_relations"].append(deprel)
else:
relations_dict[buzz]["syntax_relations"] = [deprel]
with open("before_2016_relations_raw.json", 'w', encoding='utf8') as json_file:
json.dump(relations_dict, json_file, ensure_ascii=False)
return relations_dict
def calculate_relations(rel_dict, time):
calculated_relations = recursive_dict()
relations_list = []
for buzz in rel_dict:
token_relations = rel_dict[buzz]['token_relations']
syntax_relations = rel_dict[buzz]['syntax_relations']
relation_tokens = len(token_relations)
relation_types = len(set(token_relations))
relations = len(set(syntax_relations))
relations_formula = relations * (relation_types / relation_tokens)
calculated_relations[buzz]['relation_tokens'] = relation_tokens
calculated_relations[buzz]['relation_types'] = relation_types
calculated_relations[buzz]['relations'] = relations
calculated_relations[buzz]['relations_formula'] = relations_formula
relations_list.append({"Buzzword": buzz, "Relation_Formula": relations_formula,
"Relation_Types": relation_types, "Relation_Tokens": relation_tokens,
"Relations": relations})
df = pd.DataFrame(relations_list)
df.to_csv("{}_relations_summary.csv".format(time), encoding='utf8', index=False)
return calculated_relations, relations_list
if __name__ == "__main__":
rel_dict_after_2016 = get_relations_after_2016()
calculate_relations(rel_dict_after_2016, "after_2016")
rel_dict_before_2016 = get_relations_before_2016()
calculate_relations(rel_dict_before_2016, "before_2016")