bluedistro

# A python implementation of the Banker's Algorithm in Operating Systems using
# Processes and Resources
# {
# "Author: "Biney Kingsley (bluedistro@github.io)",
# "Date": 28-10-2018
# }
'''
The Banker's algorithm is a resource allocation and deadlock avoidance algorithm
developed by Edsger Dijkstra that tests for safety by simulating the allocation of
predetermined maximum possible amounts of all resources, and then makes a "s-state"

bluedistro

imdb_dir = '../datasets/aclImdb/aclImdb'
train_dir = os.path.join(imdb_dir, 'train')
labels = list()
texts = list()

# Processing the labels of the raw IMDB data
for label_type in ['neg', 'pos']:
    dir_name = os.path.join(train_dir, label_type)
    for fname in os.listdir(dir_name):
        if fname[-4:] == '.txt':

bluedistro

# cut off reviews after 500 words
max_len = 500 
# train on 10000 samples
training_samples = 10000
 # validate on 10000 samples 
validation_samples = 10000
# consider only the top 10000 words
max_words = 10000 

# import tokenizer with the consideration for only the top 500 words

bluedistro

    # decode the words
    reverse_word_index = dict([(value, key) for (key, value) in word_index.items()])
    decoded_review = ' '.join([reverse_word_index.get(i, '?') for i in sequences[0]])

bluedistro

# model developing
text_input_layer = Input(shape=(500,))
embedding_layer = Embedding(max_words, 50)(text_input_layer)
text_layer = Conv1D(256, 3, activation='relu')(embedding_layer)
text_layer = MaxPooling1D(3)(text_layer)
text_layer = Conv1D(256, 3, activation='relu')(text_layer)
text_layer = MaxPooling1D(3)(text_layer)
text_layer = Conv1D(256, 3, activation='relu')(text_layer)
text_layer = MaxPooling1D(3)(text_layer)
text_layer = Conv1D(256, 3, activation='relu')(text_layer)

bluedistro

history = model.fit(x_train, y_train, epochs=50, batch_size=128, callbacks=callback_list,
                    validation_data=(x_val, y_val))

bluedistro

callback_list = [
    keras.callbacks.EarlyStopping(
        patience=1,
        monitor='acc',
    ),
    
    keras.callbacks.TensorBoard(
        log_dir='log_dir_m1',
        histogram_freq=1,
        embeddings_freq=1,

bluedistro

# save the tokenizer
with open(os.path.join(tokenizer_path, 'tokenizer_m1.pickle'), 'wb') as handle:
    pk.dump(tokenizer, handle, protocol=pk.HIGHEST_PROTOCOL)

bluedistro

tokenizer_path = 'tokenizer'
model_path = 'model'
model_file = os.path.join(model_path, 'movie_sentiment_m1.h5')
tokenizer_file = os.path.join(tokenizer_path, 'tokenizer_m1.pickle')
model = load_model(model_file)

# load tokenizer
with open(tokenizer_file, 'rb') as handle:
    tokenizer = pickle.load(handle)

bluedistro

def review_rating(score, decoded_review):
    if float(score) >= 0.9:
        print('Review: {}\nSentiment: Strongly Positive\nScore: {}'.format(decoded_review, score))
    elif float(score) >= 0.7 and float(score) < 0.9:
        print('Review: {}\nSentiment: Positive\nScore: {}'.format(decoded_review, score))
    elif float(score) >= 0.5 and float(score) < 0.7:
        print('Review: {}\nSentiment: Okay\nScore: {}'.format(decoded_review, score))
    else:
        print('Review: {}\nSentiment: Negative\nScore: {}'.format(decoded_review, score))
    print('\n\n')