In [1]:

    

import pandas as pd
import numpy as np
%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use('ggplot')


    

In [2]:

    

data = pd.read_csv('imdb_labelled.txt', sep='\|', engine='python', header=None)
data.columns = ['text', 'label']


    

In [3]:

    

data.head()


    

    
Out[3]:







  

    

      

      
text
      
label
    
  
  

    

      
0
      
A very, very, very slow-moving, aimless movie ...
      
0
    
    

      
1
      
Not sure who was more lost - the flat characte...
      
0
    
    

      
2
      
Attempting artiness with black & white and cle...
      
0
    
    

      
3
      
Very little music or anything to speak of.
      
0
    
    

      
4
      
The best scene in the movie was when Gerardo i...
      
1
    
  

In [4]:

    

ax = data.label.value_counts().plot(kind='bar', rot=0, title='Counts per Label')


    

In [5]:

    

import nltk
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
stopwords = set(stopwords.words('english'))
from nltk.tokenize import RegexpTokenizer

def remove_stopwords(sent):
    words = word_tokenize(sent)
    return ' '.join([w for w in words if w not in stopwords])   
    
def remove_punctuations(sent):
    regex_tok = RegexpTokenizer(r'\w+')
    words = regex_tok.tokenize(sent)
    return ' '.join(words)

def preprocess_sent(sent):
    sent = sent.lower()
    return remove_punctuations(remove_stopwords(sent))

preprocessed_text = data.text.apply(preprocess_sent)
data.loc[:, 'text'] = preprocessed_text


    

In [6]:

    

data.head()


    

    
Out[6]:







  

    

      

      
text
      
label
    
  
  

    

      
0
      
slow moving aimless movie distressed drifting ...
      
0
    
    

      
1
      
sure lost flat characters audience nearly half...
      
0
    
    

      
2
      
attempting artiness black white clever camera ...
      
0
    
    

      
3
      
little music anything speak
      
0
    
    

      
4
      
best scene movie gerardo trying find song keep...
      
1
    
  

In [7]:

    

test_df = data.groupby('label').apply(lambda g: g.sample(frac=0.2, random_state=1000))
test_df.index = test_df.index.droplevel(0)
train_df = data.drop(test_df.index, axis=0)
print("Train and test lengths: ", len(train_df), len(test_df))


    

    




Train and test lengths:  800 200

In [8]:

    

from keras.preprocessing.text import Tokenizer
from keras.preprocessing.sequence import pad_sequences

vocab_size = 40000
tokenizer = Tokenizer(lower=True, num_words=vocab_size)
tokenizer.fit_on_texts(train_df.text.values)
pad_len = 15

train_sequences = pad_sequences(tokenizer.texts_to_sequences(train_df.text.values), pad_len)
test_sequences = pad_sequences(tokenizer.texts_to_sequences(test_df.text.values), pad_len)


    

    




Using TensorFlow backend.

In [9]:

    

print("Shape of train sequences: ", train_sequences.shape)
print("Shape of test sequences : ", test_sequences.shape)


    

    




Shape of train sequences:  (800, 15)
Shape of test sequences :  (200, 15)

In [10]:

    

train_sequences[:5]


    

    
Out[10]:





array([[  0,   0,   0,   0,   0,   0,   0, 193, 243, 887,   1, 888, 889,
        244,  59],
       [  0,   0,   0,   0,   0,   0,   0, 342, 245, 501,  16, 194, 890,
        156, 891],
       [343, 126, 502,   1, 503, 894,  15, 195,  10, 196,  22, 157, 111,
        246, 895],
       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,  38,  85,
        127, 344],
       [  0,   0,   0,   0,   0,  29, 112,   1, 896, 345,  86, 346, 504,
        505, 347]])

In [66]:

    

from keras.models import Sequential
from keras.layers import Dense
from keras.layers import SimpleRNN, LSTM 
from keras.layers.embeddings import Embedding

embedding_size = 60

model = Sequential()
model.add(Embedding(input_dim=vocab_size+1, output_dim=embedding_size, mask_zero=True))
model.add(SimpleRNN(units=10, activation='tanh'))
model.add(Dense(1, activation='sigmoid'))


    

In [67]:

    

model.summary()


    

    




_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
embedding_14 (Embedding)     (None, None, 60)          2400060   
_________________________________________________________________
simple_rnn_11 (SimpleRNN)    (None, 10)                710       
_________________________________________________________________
dense_16 (Dense)             (None, 1)                 11        
=================================================================
Total params: 2,400,781
Trainable params: 2,400,781
Non-trainable params: 0
_________________________________________________________________

In [68]:

    

from keras import optimizers
sgd = optimizers.SGD(lr=0.008)
model.compile(optimizer=sgd, loss='binary_crossentropy', metrics=['accuracy'])

history = model.fit(train_sequences, train_df.label.values,
  batch_size=32,
  epochs=10,
  verbose=1,
  validation_split=0.1,
  shuffle=True)


    

    




Train on 720 samples, validate on 80 samples
Epoch 1/10
720/720 [==============================] - 3s 5ms/step - loss: 0.6982 - acc: 0.4472 - val_loss: 0.6817 - val_acc: 0.5625
Epoch 2/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6967 - acc: 0.4694 - val_loss: 0.6817 - val_acc: 0.5625
Epoch 3/10
720/720 [==============================] - 1s 2ms/step - loss: 0.6952 - acc: 0.4764 - val_loss: 0.6804 - val_acc: 0.5875
Epoch 4/10
720/720 [==============================] - 1s 2ms/step - loss: 0.6939 - acc: 0.4806 - val_loss: 0.6800 - val_acc: 0.5875
Epoch 5/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6925 - acc: 0.4931 - val_loss: 0.6802 - val_acc: 0.5750
Epoch 6/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6911 - acc: 0.5139 - val_loss: 0.6799 - val_acc: 0.5750
Epoch 7/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6898 - acc: 0.5236 - val_loss: 0.6810 - val_acc: 0.5500
Epoch 8/10
720/720 [==============================] - 1s 2ms/step - loss: 0.6885 - acc: 0.5389 - val_loss: 0.6808 - val_acc: 0.5500
Epoch 9/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6872 - acc: 0.5444 - val_loss: 0.6798 - val_acc: 0.5625
Epoch 10/10
720/720 [==============================] - 1s 1ms/step - loss: 0.6861 - acc: 0.5514 - val_loss: 0.6807 - val_acc: 0.5750

In [61]:

    

loss_per_epoch = history.history['loss']
ax = pd.Series(loss_per_epoch).plot(kind='bar', title='Loss per epoch', rot=0)


    

In [73]:

    

predictions = model.predict(test_sequences).squeeze()
predictions = np.where(predictions > 0.5, 1, 0)


    

In [75]:

    

from sklearn.metrics import classification_report
report = classification_report(predictions, test_df.label.values)
print(report)


    

    




             precision    recall  f1-score   support

          0       0.64      0.52      0.57       123
          1       0.41      0.53      0.46        77

avg / total       0.55      0.53      0.53       200