Encoder-Decoder-based-model-to-correct-tweets/code.py at master · tetsuyaAnkur/Encoder-Decoder-based-model-to-correct-tweets · GitHub

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
import random
import numpy as np
import csv
from gensim.models import Word2Vec
import theano
from keras.preprocessing.text import Tokenizer
from keras.preprocessing.text import text_to_word_sequence
from keras.models import Sequential
from keras.layers import LSTM
from keras.layers import Dense
from keras.layers import TimeDistributed
from keras.layers import RepeatVector

# this function returns one hot vector
def one_hot(x):
	y = np.zeros((3501))
	if x in output.keys():
		y[output[x]]=1
	else:
		y[3500]=1
	return y

# preprocessing
input_data=[]
output_data=[]
words=[]
with open('consolidated.csv') as csvfile:
	reader=csv.DictReader(csvfile)
	for row in reader:
		if int(row['valid'])==1:
			input_data.append(row['tweet'])
			output_data.append(row['corrected'])
			words.append([row['tweet'],row['corrected']])

input_data2=[]
for y in input_data:
	input_data2.append(text_to_word_sequence(y))

output_data2=[]
for i in output_data:
	output_data2.append(text_to_word_sequence(i))

total=[]
for x in output_data2:
	for y in x:
		total.append(y)
s = set(total)

a = []
for i in s:
	c = total.count(i)
	x = [c,i]
	a.append(x)

a.sort(reverse=True)

output = {}
for i in range(0,3500):
	output[a[i][1]]=i;

output['__unk__']=3500
data=random.sample(words,12000)
data_test=data[:int(0.0400*12000)]
data_train=data[int(0.0400*12000):]
data_train_corrected=[]
data_train_tweets=[]
data_test_corrected=[]
data_test_tweets=[]

for x in data_test:
	data_test_corrected.append(text_to_word_sequence(x[1]))
	data_test_tweets.append(text_to_word_sequence(x[0]))

for x in data_train:
	data_train_corrected.append(text_to_word_sequence(x[1]))
	data_train_tweets.append(text_to_word_sequence(x[0]))

one_hots=[]
lenk=42
for x in data_train_corrected:
	temp=[]
	for y in x:
		temp.append(np.asarray(one_hot(y)))
	k=lenk-len(temp)
	for i in range(k):
		temp.append(np.zeros((3501)))
	one_hots.append(np.asarray(temp))

one_hots=np.asarray(one_hots)

one_hots_test=[]
for x in data_test_corrected:
	temp=[]
	for y in x:
		temp.append(np.asarray(one_hot(y)))
	k=lenk-len(temp)
	for i in range(k):
		temp.append(np.zeros((3501)))
	one_hots_test.append(np.asarray(temp))

one_hots_test=np.asarray(one_hots_test)

model = Word2Vec(input_data2, min_count=1, workers=2,size=100)

inputs=[]
for x in data_train_tweets:
	temp=[]
	for y in x:
		temp.append(model.wv[y])
	k=lenk-len(temp)
	for i in range(k):
		temp.append(np.zeros((100)))
	inputs.append(np.asarray(temp))

inputs=np.asarray(inputs)

inputs_test=[]
for x in data_test_tweets:
	temp=[]
	for y in x:
		temp.append(model.wv[y])
	k=lenk-len(temp)
	for i in range(k):
		temp.append(np.zeros((100)))
	inputs_test.append(np.asarray(temp))

inputs_test=np.asarray(inputs_test)

# training
n_features = 100
n_timesteps_in = 42
n_timesteps_out = 2
model = Sequential()
model.add(LSTM(150, input_shape=(n_timesteps_in, n_features)))
model.add(RepeatVector(n_timesteps_in))
model.add(LSTM(150, return_sequences=True))
model.add(TimeDistributed(Dense(3501, activation='softmax')))
model.compile(loss='categorical_crossentropy', optimizer='adamax', metrics=['acc'])
model.fit(inputs, one_hots, batch_size=80, validation_data=(inputs_test,one_hots_test),epochs=1)