使用 全连接神经网络 和 词袋模型 进行文本分类的example

# -*- coding: utf-8 -*-
import jieba
import tensorflow as tf

def prepareTestData():
f = open("test.txt",'r')
Y = []
sentenceList = []
while True:
line = f.readline()
if len(line) == 0:
break
temp = str.split(line," ")
line = temp[0]
y = temp[1]
Y.append(y)
words = jieba.lcut(line)
#  totalWordList.extend(words)
sentenceList.append(words)

sentenceByNumberList = []

for i in range(0,len(sentenceList)):
sentenceByNumber = []
for k in range(0,len(totalWordList)):
sentenceByNumber.append(0)
for j in range(0,len(sentenceList[i])):
if sentenceList[i][j] in totalWordList:
index = totalWordList.index(sentenceList[i][j])
sentenceByNumber[index]=sentenceByNumber[index]+1
sentenceByNumberList.append(sentenceByNumber)

f.close()

Y_feed = []
for i in range(0,len(Y)):
temp = [0]*n_classes
temp[int(Y[i])-1]=1
Y_feed.append(temp)
return sentenceByNumberList,Y_feed

#下面是BOW的事
jieba.initialize()

jieba.add_word("看电视",0)
jieba.add_word("看",100,"v")
jieba.add_word("电视",100,"n")
totalWordList = []
sentenceList = []
Y = []
f = open("train.txt",'r')
while True:
line = f.readline()
if len(line) == 0:
break
temp = str.split(line," ")
line = temp[0]
y = temp[1]
Y.append(y)
words = jieba.lcut(line)
totalWordList.extend(words)
sentenceList.append(words)
totalWordList = list(set(totalWordList))

#totalWordList.remove("\n")

sentenceByNumberList = []

for i in range(0,len(sentenceList)):
sentenceByNumber = []
for k in range(0,len(totalWordList)):
sentenceByNumber.append(0)
for j in range(0,len(sentenceList[i])):
if sentenceList[i][j] in totalWordList:
index = totalWordList.index(sentenceList[i][j])
sentenceByNumber[index]=sentenceByNumber[index]+1
sentenceByNumberList.append(sentenceByNumber)

f.close()

Y_feed = []
for i in range(0,len(Y)):
temp = [0,0,0,0,0]
temp[int(Y[i])-1]=1
Y_feed.append(temp)

#下面是tensorflow的事
learning_rate = 0.001
training_epochs = 15
batch_size = 100
display_step = 1

n_hidden_1 = 256
n_hidden_2 = 256
n_input = len(totalWordList)
n_classes = len(set(Y))

x = tf.placeholder("float", [None, n_input])
y = tf.placeholder("float", [None, n_classes])

def multilayer_perceptron(x, weights, biases):
layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
layer_1 = tf.nn.relu(layer_1)
layer_2 = tf.add(tf.matmul(layer_1, weights['h2']), biases['b2'])
layer_2 = tf.nn.relu(layer_2)
out_layer = tf.matmul(layer_2, weights['out']) + biases['out']
return out_layer

weights = {
'h1': tf.Variable(tf.random_normal([n_input, n_hidden_1])),
'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
'out': tf.Variable(tf.random_normal([n_hidden_2, n_classes]))
}
biases = {
'b1': tf.Variable(tf.random_normal([n_hidden_1])),
'b2': tf.Variable(tf.random_normal([n_hidden_2])),
'out': tf.Variable(tf.random_normal([n_classes]))
}

pred = multilayer_perceptron(x, weights, biases)

cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(pred, y))
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)

init = tf.initialize_all_variables()

with tf.Session() as sess:
sess.run(init)

for epoch in range(training_epochs):
avg_cost = 0.
total_batch = 10
for i in range(total_batch):
batch_x = sentenceByNumberList
batch_y = Y_feed
_, c = sess.run([optimizer, cost], feed_dict={x: batch_x,
y: batch_y})
# 计算平均loss
avg_cost += c / total_batch
if epoch % display_step == 0:
print "Epoch:", '%04d' % (epoch+1), "cost=", \
"{:.9f}".format(avg_cost)
print "Optimization Finished!"

# 测试模型，分别用原数据和测试数据
pred = multilayer_perceptron(x, weights, biases)
correct_prediction = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
# 计算Accuracy
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
print "Accuracy:", accuracy.eval({x: sentenceByNumberList, y: Y_feed})
testX,testY = prepareTestData()
print "Accuracy:", accuracy.eval({x: testX, y: testY})

给我播放故事 5
我想听故事 5

我想看电视 1
我想听音乐 2
我想打电话 3
我想发消息 4
我想听故事 5
播放音乐 2
播放故事 5
给我放故事 5