forked from tuandat95cbn/20162_DataMining_SMS_challenge
-
Notifications
You must be signed in to change notification settings - Fork 0
/
sms-word2vec.py
284 lines (225 loc) · 8.53 KB
/
sms-word2vec.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
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
from __future__ import print_function
import collections
import math
import numpy as np
#import os
import random
import tensorflow as tf
#import zipfile
#from matplotlib import pylab
#from six.moves import range
#from six.moves.urllib.request import urlretrieve
from sklearn.manifold import TSNE
from scipy import spatial
import operator
vocabulary_size = 0
#read file and return all words in file
def read_data(filename) :
f = open(filename, 'r')
t = f.read().splitlines()
f.close()
data = []
for line in t :
words = line.split(" ")
for i in range(1,len(words)) :
data.append(words[i])
global vocabulary_size
vocabulary_size = len(set(data))
return data
words = read_data('train.txt')
print(set(words))
#build dictionary and map word to index in dictonary
def build_dataset(words) :
#order word by appearance
count = []
count.extend(collections.Counter(words).most_common(vocabulary_size))
#{word : index of word in dict}
dictionary = dict()
for word, _ in count :
dictionary[word] = len(dictionary)
#dictionary's index of per word in words
data = list()
for word in words :
index = dictionary[word]
data.append(index)
#reverse dictionary : {index : word}
reverse_dictionary = dict(zip(dictionary.values(),dictionary.keys()))
return data, count, dictionary, reverse_dictionary
data, count, dictionary, reverse_dictionary = build_dataset(words)
del words
print("len(data) = ",len(data))
data_index = 0
def generate_batch(batch_size, num_skips, skip_window) :
global data_index
assert batch_size % num_skips == 0
assert num_skips <= 2 * skip_window
batch = np.ndarray(shape=(batch_size),dtype = np.int32)
labels = np.ndarray(shape=(batch_size,1),dtype = np.int32)
span = 2 * skip_window + 1 #[skip_window target skip_window]
buffer = collections.deque(maxlen = span)
for _ in range(span) :
buffer.append(data[data_index])
data_index = (data_index + 1) % len(data)
for i in range(batch_size // num_skips) :
target = skip_window
targets_to_avoid = [skip_window]
for j in range(num_skips) :
while target in targets_to_avoid :
target = random.randint(0,span -1)
targets_to_avoid.append(target)
batch[i * num_skips + j] = buffer[skip_window]
labels[i * num_skips + j] = buffer[target]
buffer.append(data[data_index])
data_index = (data_index + 1) % len(data)
return batch, labels
batch_size = 128
embedding_size = 100 # Dimension of the embedding vector.
skip_window = 1 # How many words to consider left and right.
num_skips = 2 # How many times to reuse an input to generate a label.
# We pick a random validation set to sample nearest neighbors. here we limit the
# validation samples to the words that have a low numeric ID, which by
# construction are also the most frequent.
valid_size = 16 # Random set of words to evaluate similarity on.
valid_window = 100 # Only pick dev samples in the head of the distribution.
valid_examples = np.array(random.sample(range(valid_window), valid_size))
num_sampled = 64 # Number of negative examples to sample.
graph = tf.Graph()
with graph.as_default(), tf.device('/cpu:0') :
#tf.placholder : data
train_dataset = tf.placeholder(tf.int32, shape=[batch_size])
train_labels = tf.placeholder(tf.int32, shape=[batch_size,1])
valid_dataset = tf.constant(valid_examples, dtype=tf.int32)
#tf.Variable : variable like embedding vectors
#generate matrix embedding with value in [-1,1]
embeddings = tf.Variable(tf.random_uniform([vocabulary_size,embedding_size],-1.0,1.0))
softmax_weights = tf.Variable(tf.truncated_normal([vocabulary_size,embedding_size],stddev=1.0/math.sqrt(embedding_size)))
softmax_biases = tf.Variable(tf.zeros([vocabulary_size]))
embed = tf.nn.embedding_lookup(embeddings, train_dataset)
loss = tf.reduce_mean(
tf.nn.sampled_softmax_loss(weights=softmax_weights, biases=softmax_biases, inputs=embed,
labels=train_labels, num_sampled=num_sampled, num_classes=vocabulary_size))
optimizer = tf.train.AdagradOptimizer(1.0).minimize(loss)
norm = tf.sqrt(tf.reduce_sum(tf.square(embeddings), 1, keep_dims=True))
normalized_embeddings = embeddings / norm
valid_embeddings = tf.nn.embedding_lookup(normalized_embeddings, valid_dataset)
similarity = tf.matmul(valid_embeddings, tf.transpose(normalized_embeddings))
num_steps = 100001
final_embeddings = np.array([])
with tf.Session(graph=graph) as session:
tf.global_variables_initializer().run()
print('Initialized')
average_loss = 0
for step in range(num_steps):
batch_data, batch_labels = generate_batch(batch_size, num_skips, skip_window)
feed_dict = {train_dataset : batch_data, train_labels : batch_labels}
_, l = session.run([optimizer, loss], feed_dict=feed_dict)
average_loss += l
if step % 2000 == 0:
if step > 0:
average_loss = average_loss / 2000
print('Average loss at step %d: %f' % (step, average_loss))
average_loss = 0
if step % 10000 == 0:
sim = similarity.eval()
for i in range(valid_size):
valid_word = reverse_dictionary[valid_examples[i]]
top_k = 8 # number of nearest neighbors
nearest = (-sim[i, :]).argsort()[1:top_k+1]
log = 'Nearest to %s:' % valid_word
for k in range(top_k):
close_word = reverse_dictionary[nearest[k]]
log = '%s %s,' % (log, close_word)
print(log)
global final_embeddings
final_embeddings = normalized_embeddings.eval()
#print("final_embedding size",len(final_embeddings))
#print("final_embeddings[0] = ",final_embeddings[0])
#return list sms that each sms contains list word embedded vector
def convert_sms_to_vector(filename,type) :
f = open(filename,'r')
t = f.read().splitlines()
f.close()
list_sms_vector = []
for line in t :
sms_vector = []
words = line.split(" ")
if type == 'trainning' :
sms_vector.append([words[0]])
for i in range(1,len(words)) :
if words[i] in dictionary :
index_in_dict = dictionary[words[i]]
#index_in_data = data.index(index_in_dict)
word_vector = final_embeddings[index_in_dict].tolist()
else :
word_vector = [0] * embedding_size
sms_vector.append(word_vector)
list_sms_vector.append(sms_vector)
return list_sms_vector
#list_sms_vector = convert_sms_to_vector('train.txt','trainning')
#print("sms[0] = ",list_sms_vector[0])
def convert_all_sms_same_length(list_sms_vector) :
maxlen = -1
for i in range(len(list_sms_vector)) :
if maxlen < len(list_sms_vector[i]) :
maxlen = len(list_sms_vector[i])
for i in range(len(list_sms_vector)) :
if len(list_sms_vector[i]) < maxlen :
for _ in range(len(list_sms_vector[i]),maxlen) :
list_sms_vector[i].append([0] * embedding_size)
return list_sms_vector
def split_set(input_set,split) :
trainning_set = []
test_set = []
for i in range(len(input_set)) :
if random.random() < split :
trainning_set.append(input_set[i])
else :
test_set.append(input_set[i])
return trainning_set, test_set
def cosine_distance(instance1, instance2) :
distance = 0
for i in range(len(instance1)) :
diff = 0
for j in range(len(instance1[i])) :
diff += math.pow(instance1[i][j] - instance2[i][j],2)
distance += math.sqrt(diff)/float(len(instance1[i]))
return distance/float(len(instance1))
#get k neighbors neariest test_instance
def getNeighbors(training_set,test_instance,k) :
distances = []
#print("training_set[0][0]",training_set[0][0])
for train_instance in training_set :
dist = cosine_distance(train_instance[1:], test_instance)
distances.append((train_instance,dist))
distances.sort(key=operator.itemgetter(1))
neighbors = []
for i in range(k) :
neighbors.append(distances[i][0])
return neighbors
#get max label in neighbors
def getResponse(neighbors) :
classVotes = {}
for neighbor in neighbors :
response = neighbor[0][0]
if response in classVotes :
classVotes[response] += 1
else:
classVotes[response] = 1
sortedVotes = sorted(classVotes.iteritems(),key=operator.itemgetter(1),reverse=True)
return sortedVotes[0][0]
def main() :
split = 0.8
k = 10
list_sms_vector_trainning = convert_sms_to_vector('train.txt','trainning')
#print("list_sms_vector_trainning[0] = ",list_sms_vector_trainning[0])
list_sms_vector_trainning = convert_all_sms_same_length(list_sms_vector_trainning)
training_set, test_set = split_set(list_sms_vector_trainning,split)
instance_correct = 0
for test_item in test_set :
neighbors = getNeighbors(training_set,test_item[1:],k)
label = getResponse(neighbors)
if label == test_item[0][0] :
instance_correct += 1
accurancy = (instance_correct/float(len(test_set)))*100
print("accurancy = ",accurancy)
main()