keras—多层感知器MLP—IMDb情感分析

import urllib.request
import os
import tarfile
from keras.datasets import imdb
from keras.preprocessing import sequence
from keras.preprocessing.text import Tokenizer
import re
def rm_tags(text):
    re_tag=re.compile(r'<[^>]+>')
    return re_tag.sub('',text)
def read_files(filetype):
    path="C:/Users/admin/.keras/aclImdb/"
    file_list=[]
    positive_path=path+filetype+"/pos/"
    for f in os.listdir(positive_path):
        file_list+=[positive_path+f]
    negative_path=path+filetype+"/pos/"
    for f in os.listdir(negative_path):
        file_list+=[negative_path+f]
    print('read',filetype,'files:',len(file_list))
    all_labels=([1]*12500+[0]*12500)
    all_texts=[]
    for fi in file_list:
        with open(fi,encoding='utf8') as file_input:
            all_texts+=[rm_tags(" ".join(file_input.readlines()))]
    return all_labels,all_texts
y_train,train_text=read_files("train")
y_test,test_text=read_files("test")
print(train_text[0])
print(y_train[0])
token=Tokenizer(num_words=2000)
token.fit_on_texts(train_text)
print(token.document_count)
print(token.word_index)
x_train_seq=token.texts_to_sequences(train_text)
x_test_seq=token.texts_to_sequences(test_text)
print(train_text[0])
print(x_train_seq[0])
x_train=sequence.pad_sequences(x_train_seq,maxlen=100)
x_test=sequence.pad_sequences(x_test_seq,maxlen=100)
print('before pad_sequences lenfth=',len(x_train_seq[113]))
print(x_train_seq[113])
print('after pad_sequences lenfth=',len(x_train[113]))
print(x_train[113])
from keras.models import Sequential
from keras.layers import Dense,Dropout,Flatten,Activation
from keras.layers.embeddings import Embedding
model=Sequential()
model.add(Embedding(output_dim=32,
                 input_dim=2000,
                 input_length=100))
model.add(Dropout(0.2))
#model.add(SimpleRNN(units=16))
model.add(Flatten())
model.add(Dense(units=256,
                activation='relu'))
model.add(Dropout(0.35))
model.add(Dense(units=1,
                activation='sigmoid'))
print(model.summary())
model.compile(loss='binary_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])
train_history=model.fit(x=x_train,y=y_train,batch_size=100,
                        epochs=10,verbose=2,
                        validation_split=0.2)
scores=model.evaluate(x_test,y_test,verbose=1)
print('accuracy',scores[1])
predict=model.predict_classes(x_test)
print("prediction[:10]",predict[:10])
predict_classes=predict.reshape(-1)
print(predict_classes[:10])
SentimentDict = {1: '正面的', 0: '负面的'}
def display_test_Sentiment(i):
    print(test_text[i])
    print('label真实值:', SentimentDict[y_test[i]],
          '预测结果:', SentimentDict[predict_classes[i]])
display_test_Sentiment(12502)
input_text='''
I saw this film with my 6-year-old a couple weeks ago. While there's plenty about which to gripe, here's one of 
my biggest problems: I can't stand this constant CGI-heavy everything-must-be-a-sequel-or- a- remake era of film
making. It's making movie makers lazy.
'''
input_seq=token.texts_to_sequences([input_text])
len(input_seq[0])
print(input_seq[0])
pad_input_seq=sequence.pad_sequences(input_seq,maxlen=100)
len(pad_input_seq[0])
print(pad_input_seq[0])
predict_result=model.predict_classes(pad_input_seq)
print(predict_result)
print(predict_result[0][0])
print(SentimentDict[predict_result[0][0]])
def predict_review(input_text):
    input_seq=token.texts_to_sequences([input_text])
    pad_input_seq=sequence.pad_sequences(input_seq,maxlen=100)
    predict_result=model.predict_classes(pad_input_seq)
    print(SentimentDict[predict_result[0][0]])

predict_review('''
They poured on the whole "LeFou is gay" thing a bit thick for my taste. It was the only thing that added levity to the movie (despite how much fun it should have been already), but it seemed a bit cheap. I'm not going to apologize for wanting more for my LGBTQ characters than to be just the comic relief.
''')

验证的准确率为0问题待解决