-
Notifications
You must be signed in to change notification settings - Fork 0
/
toolsIT_bert.py
265 lines (225 loc) · 9.79 KB
/
toolsIT_bert.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
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
import numpy as np
import pyfreeling
import os
import json
emoji = {':-)': 'sorridi', '(-:': 'sorridi', ':)': 'sorridi', '(:': 'sorridi', '=]': 'sorridi', ':]': 'sorridi',
':d': 'ridere', '8d': 'ridere', 'xd': 'ridere', '=d': 'ridere', ':-d': 'ridere',
':D': 'ridere', '8D': 'ridere', 'xD': 'ridere', '=D': 'ridere', ':-D': 'ridere',
'x-d': 'ridere', ':(': 'triste', '):': 'triste', ':c': 'triste', ':[': 'triste', ':\'(': 'pianto',
':,( ': 'pianto', 'x-D': 'ridere', ':"(': 'pianto', ':((': 'pianto', ':|': 'neutrale',
'=_=': 'neutrale', '-_-': 'neutrale', ':-\\': 'neutrale', ':O': 'neutrale', ':-!': 'neutrale'}
AGE_CLASSES = ['0-19', '20-29', '30-39', '40-49', '50-100']
TOPIC_CLASSES = ['ANIME', 'AUTO-MOTO', 'BIKES', 'CELEBRITIES', 'ENTERTAINMENT', 'MEDICINE-AESTHETICS',
'METAL-DETECTING', 'NATURE', 'SMOKE', 'SPORTS', 'TECHNOLOGY']
def load_stylistic_features(STYLE_PATH):
with open(STYLE_PATH) as F:
data = json.load(F)
STYF = [np.fromstring(data['training']['txt'][str(i+1)]['vec'], 'f', sep=',') for i in range(len(data['training']['txt']))]
return np.array(STYF)
def load_data(data_path):
postxuser = []
print(data_path)
XTree = open(data_path, mode='rU')
users = []
band = False
for line in XTree:
usr = line.split(' ')
if usr[0] == "<user":
users.append({'post': '', 'id': int(usr[1][4:len(usr[1]) - 1]), 'age': usr[3][len("topic"):len(usr[3]) - 1],
'gender': usr[4][len("gender= "):len(usr[4]) - 3],
'topic': usr[2][len("topic= "):len(usr[2]) - 1]})
postxuser.append(0)
if band:
users[len(users) - 1]['post'] += " < post > "
if line[:len(line) - 1] == '':
users[len(users) - 1]['post'] += " < vuoto > ."
else:
tmp = line[:len(line) - 1]
for k in emoji.keys():
while True:
z = str.find(tmp, k)
if z == -1:
break
tmp1 = tmp[z + len(k):]
tmp = tmp[:z] + '< emoticon > ' + emoji[k] + ' '
tmp += tmp1
users[len(users) - 1]['post'] += tmp
if users[len(users) - 1]['post'][len(users[len(users) - 1]['post']) - 1] not in {'?', '!', '.', '...'}:
users[len(users) - 1]['post'] += '.'
band = False
if usr[0][:len(usr[0]) - 1] == "<post>":
band = True
postxuser[len(postxuser) - 1] += 1
return users, postxuser
def Preprocess(users):
print('Freeling Tokenizing ', '0 %\r', end="")
if "FREELINGDIR" not in os.environ:
os.environ["FREELINGDIR"] = "/usr/local"
DATA = os.environ["FREELINGDIR"] + "/share/freeling/"
pyfreeling.util_init_locale("default")
LANG = "it"
op = pyfreeling.maco_options(LANG)
op.set_data_files("",
DATA + "common/punct.dat",
DATA + LANG + "/dicc.src",
DATA + LANG + "/afixos.dat",
"",
DATA + LANG + "/locucions.dat",
DATA + LANG + "/np.dat",
"", # DATA + LANG + "/quantities.dat",
DATA + LANG + "/probabilitats.dat")
tk = pyfreeling.tokenizer(DATA + LANG + "/tokenizer.dat")
sp = pyfreeling.splitter(DATA + LANG + "/splitter.dat")
mf = pyfreeling.maco(op)
mf.set_active_options(False, True, True, True, # select which among created
True, True, True, True, # submodules are to be used.
True, True, False, True) # default: all created submodules are used
tg = pyfreeling.hmm_tagger(DATA + LANG + "/tagger.dat", True, 2)
sen = pyfreeling.senses(DATA + LANG + "/senses.dat")
done = 0
perc = 0
top = len(users)
cont = 0
for i in range(len(users)):
done += 1
z = done / top
z *= 100
z = int(z)
if z - perc >= 1:
perc = z
print('Freeling Tokenizing ', str(perc) + ' %\r', end="")
x = users[i]['post']
l = tk.tokenize(x)
# sid=sp.open_session();
ls = sp.split(l)
ls = mf.analyze(ls)
ls = tg.analyze(ls)
ls = sen.analyze(ls)
listsent = []
for s in ls:
cont += 1
ora = s.get_words()
tmp = []
for k in range(len(ora)):
if ora[k].get_tag() == 'W':
tmp.append( 'data' )
elif ora[k].get_tag() == 'Z':
tmp.append( 'numero' )
else:
tmp.append(ora[k].get_form().lower())
# print(ora[k].get_form().lower(), ora[k].get_tag())
listsent.append(tmp)
users[i]['post'] = np.array(listsent)
print('Freeling Tokenizing ok ')
return users
def load_for_age(DATA_PATH, mode):
def load_data_AGE(data_path):
postxuser = []
print(data_path)
XTree = open(data_path, mode='rU')
users = []
band = False
for line in XTree:
usr = line.split(' ')
if usr[0] == "<user":
users.append({'post': '', 'id': int(usr[1][4:len(usr[1]) - 1]), 'age': usr[3][len("topic"):len(usr[3]) - 1],
'gender': usr[4][len("gender= "):len(usr[4]) - 3],
'topic': usr[2][len("topic= "):len(usr[2]) - 1]})
postxuser.append(0)
if band:
if line[:len(line) - 1] == '':
band = False
continue
else:
tmp = line[:len(line) - 1]
for k in emoji.keys():
while True:
z = str.find(tmp, k)
if z == -1:
break
tmp1 = tmp[z + len(k):]
tmp = tmp[:z] + '< emoticon > '
tmp += tmp1
users[len(users) - 1]['post'] += tmp
if users[len(users) - 1]['post'][len(users[len(users) - 1]['post']) - 1] not in {'?', '!', '.', '...'}:
users[len(users) - 1]['post'] += '.'
users[len(users) - 1]['post'] += ' '
band = False
if usr[0][:len(usr[0]) - 1] == "<post>":
band = True
postxuser[len(postxuser) - 1] += 1
return users, postxuser
def Preprocess_AGE(users):
print('Freeling Tokenizing ', '0 %\r', end="")
if "FREELINGDIR" not in os.environ:
os.environ["FREELINGDIR"] = "/usr/local"
DATA = os.environ["FREELINGDIR"] + "/share/freeling/"
pyfreeling.util_init_locale("default")
LANG = "it"
op = pyfreeling.maco_options(LANG)
op.set_data_files("",
DATA + "common/punct.dat",
DATA + LANG + "/dicc.src",
DATA + LANG + "/afixos.dat",
"",
DATA + LANG + "/locucions.dat",
DATA + LANG + "/np.dat",
"", # DATA + LANG + "/quantities.dat",
DATA + LANG + "/probabilitats.dat")
tk = pyfreeling.tokenizer(DATA + LANG + "/tokenizer.dat")
sp = pyfreeling.splitter(DATA + LANG + "/splitter.dat")
mf = pyfreeling.maco(op)
mf.set_active_options(False, True, True, True, # select which among created
True, True, True, True, # submodules are to be used.
True, True, False, True) # default: all created submodules are used
tg = pyfreeling.hmm_tagger(DATA + LANG + "/tagger.dat", True, 2)
sen = pyfreeling.senses(DATA + LANG + "/senses.dat")
done = 0
perc = 0
top = len(users)
cont = 0
for i in range(len(users)):
done += 1
z = done / top
z *= 100
z = int(z)
if z - perc >= 1:
perc = z
print('Freeling Tokenizing ', str(perc) + ' %\r', end="")
x = users[i]['post']
l = tk.tokenize(x)
# sid=sp.open_session();
ls = sp.split(l)
ls = mf.analyze(ls)
ls = tg.analyze(ls)
ls = sen.analyze(ls)
listsent = []
for s in ls:
cont += 1
ora = s.get_words()
tmp = []
for k in range(len(ora)):
x = ora[k].get_form().lower()
if ora[k].get_tag() == 'W':
listsent.append('data')
elif ora[k].get_tag() == 'Z':
listsent.append('numero')
else: listsent.append(x)
users[i]['post'] = np.array(listsent)
print('Freeling Tokenizing ok ')
return users
profiles, postxuser = load_data_AGE(DATA_PATH)
profiles = Preprocess_AGE(profiles)
TOP_SENT_LENGHT = 348
to_encode = np.array(np.zeros((len(profiles), TOP_SENT_LENGHT)), dtype=np.str_)
for i in range(len(profiles)):
for j in range(int(min(TOP_SENT_LENGHT, len(profiles[i]['post'])))):
to_encode[i][j] = profiles[i]['post'][j]
labels = np.zeros((len(to_encode), 3))
if mode != 'test':
for i in range(len(profiles)):
labels[i][0] = AGE_CLASSES.index(profiles[i]['age'])
labels[i][1] = TOPIC_CLASSES.index(profiles[i]['topic'])
if profiles[i]['gender'] == 'F':
labels[i][2] = 1
return to_encode, labels