python - 预测情绪分析时出错 Tensorflow NLTK

Question

我正在使用 Tensorflow 框架学习情绪分析。

我正在学习来自 pythonprogramming_tutorial(create_feature_sets_and_labels) 的教程和 pythonprogramming_tutorial(train_test)

在create_sentiment_featuresets.py(第一个链接)中，我只添加了一个方法来获取词典并修改了给定的代码sentiment_demo.py(第二个链接) ) 测试给定输入字符串的情绪。

create_sentiment_featuresets.py

import nltk
from nltk.tokenize import word_tokenize
import numpy as np
import random
import pickle
from collections import Counter
from nltk.stem import WordNetLemmatizer

lemmatizer = WordNetLemmatizer()
hm_lines = 100000
def create_lexicon(pos, neg):

    lexicon = []
    with open(pos, 'r') as f:
        contents = f.readlines()            # readline vs strip
        for l in contents[:len(contents)]:
            l= l.decode('utf-8')
            all_words = word_tokenize(l)
            lexicon += list(all_words)

    f.close()

    with open(neg, 'r') as f:
        contents = f.readlines()            # readline vs strip
        for l in contents[:len(contents)]:
            l= l.decode('utf-8')
            all_words = word_tokenize(l)
            lexicon += list(all_words)

    f.close()

    lexicon = [lemmatizer.lemmatize(i) for i in lexicon]
    w_counts = Counter(lexicon)
    #print(len(w_counts))
    l2 = []
    for w in w_counts:
        if 1000 > w_counts[w] > 50:
            l2.append(w)
    #print(len(l2))
    #print(l2)
    print("Lexicon length create_lexicon: ",len(lexicon))

    return l2

def sample_handling(sample, lexicon, classification):

    featureset = []
    print("Lexicon length Sample handling: ",len(lexicon))
    with open(sample, 'r') as f:
        contents = f.readlines()
        for l in contents[:len(contents)]:
            l= l.decode('utf-8')
            current_words = word_tokenize(l.lower())
            current_words= [lemmatizer.lemmatize(i) for i in current_words]

            features = np.zeros(len(lexicon))
            for word in current_words:
                if word.lower() in lexicon:
                    index_value = lexicon.index(word.lower())
                    features[index_value] +=1

            features = list(features)
            featureset.append([features, classification])
    f.close()
    print("Feature SET------")
    print(len(featureset))
    return featureset

def create_feature_sets_and_labels(pos, neg, test_size = 0.1):
    global m_lexicon
    m_lexicon = create_lexicon(pos, neg)
    features = []
    features += sample_handling(pos, m_lexicon, [1,0])
    features += sample_handling(neg, m_lexicon, [0,1])

    random.shuffle(features)
    features = np.array(features)

    testing_size = int(test_size * len(features))

    train_x = list(features[:,0][:-testing_size])
    #print("TRAIN_X", train_x)
    train_y = list(features[:,1][:-testing_size])
    #print("TRAIN_Y", train_y)
    test_x = list(features[:,0][-testing_size:])
    test_y = list(features[:,1][-testing_size:])

    return train_x, train_y, test_x, test_y

def get_lexicon():
    global m_lexicon
    return m_lexicon

为了训练和测试，我使用了第一个链接中给出的 pos.txt 和 neg.txt。文件分别包含5000个正面和负面的句子

下面是我的sentiment_demo.py:

from create_sentiment_featuresets import create_feature_sets_and_labels
from create_sentiment_featuresets import get_lexicon

import tensorflow as tf
import pickle
import numpy as np

# extras for testing
from nltk.tokenize import word_tokenize 
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
#- end extras


train_x, train_y, test_x, test_y = create_feature_sets_and_labels('pos.txt', 'neg.txt')

n_nodes_hl1 = 1500
n_nodes_hl2 = 1500
n_nodes_hl3 = 1500

n_classes = 2
batch_size = 100
hm_epochs = 5

x = tf.placeholder('float')
y = tf.placeholder('float')

hidden_1_layer = {'f_fum': n_nodes_hl1,
                'weight': tf.Variable(tf.random_normal([len(train_x[0]), n_nodes_hl1])),
                'bias': tf.Variable(tf.random_normal([n_nodes_hl1]))}
hidden_2_layer = {'f_fum': n_nodes_hl2,
                'weight': tf.Variable(tf.random_normal([n_nodes_hl1, n_nodes_hl2])),
                'bias': tf.Variable(tf.random_normal([n_nodes_hl2]))}
hidden_3_layer = {'f_fum': n_nodes_hl3,
                'weight': tf.Variable(tf.random_normal([n_nodes_hl2, n_nodes_hl3])),
                'bias': tf.Variable(tf.random_normal([n_nodes_hl3]))}
output_layer = {'f_fum': None,
                'weight': tf.Variable(tf.random_normal([n_nodes_hl3, n_classes])),
                'bias': tf.Variable(tf.random_normal([n_classes]))}


def nueral_network_model(data):

    l1 = tf.add(tf.matmul(data, hidden_1_layer['weight']), hidden_1_layer['bias'])
    l1 = tf.nn.relu(l1)

    l2 = tf.add(tf.matmul(l1, hidden_2_layer['weight']), hidden_2_layer['bias'])
    l2 = tf.nn.relu(l2)

    l3 = tf.add(tf.matmul(l2, hidden_3_layer['weight']), hidden_3_layer['bias'])
    l3 = tf.nn.relu(l3)

    output = tf.matmul(l3, output_layer['weight']) + output_layer['bias']

    return output

def train_neural_network(x):
    prediction = nueral_network_model(x)
    cost = tf.reduce_mean( tf.nn.softmax_cross_entropy_with_logits(logits= prediction, labels= y))
    optimizer = tf.train.AdamOptimizer(learning_rate= 0.001).minimize(cost)



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

        for epoch in range(hm_epochs):
            epoch_loss = 0
            i = 0
            while i < len(train_x):
                start = i
                end = i+ batch_size
                batch_x = np.array(train_x[start: end])
                batch_y = np.array(train_y[start: end])

                _, c = sess.run([optimizer, cost], feed_dict= {x: batch_x, y: batch_y})
                epoch_loss += c
                i+= batch_size
            print('Epoch', epoch+ 1, 'completed out of ', hm_epochs, 'loss:', epoch_loss)

        correct= tf.equal(tf.argmax(prediction, 1), tf.argmax(y, 1))
        accuracy = tf.reduce_mean(tf.cast(correct, 'float'))

        print('Accuracy:', accuracy.eval({x:test_x, y:test_y}))


        # testing ------Trying to predict the sentiment for an input string--------
        m_lexicon= get_lexicon()
        print('Lexicon length: ',len(m_lexicon))

        input_data= "He is an idiot"

        current_words= word_tokenize(input_data.lower())
        current_words = [lemmatizer.lemmatize(i) for i in current_words]
        features = np.zeros(len(m_lexicon))

        for word in current_words:
            if word.lower() in m_lexicon:
                index_value = m_lexicon.index(word.lower())
                features[index_value] +=1

        features = np.array(list(features))
        print('features length: ',len(features))
        result = sess.run(tf.argmax(prediction.eval(feed_dict={x:features}), 1))
        print('RESULT: ', result)
        if result[0] == 0:
            print('Positive: ', input_data)
        elif result[0] == 1:
            print('Negative: ', input_data)


train_neural_network(x)

程序一直运行到 epoch loss 的打印，之后出现以下错误:

('Epoch', 1, 'completed out of ', 5, 'loss:', 1289814.4057617188)
('Epoch', 2, 'completed out of ', 5, 'loss:', 457882.97705078125)
('Epoch', 3, 'completed out of ', 5, 'loss:', 243073.83074951172)
('Epoch', 4, 'completed out of ', 5, 'loss:', 245525.22399902344)
('Epoch', 5, 'completed out of ', 5, 'loss:', 233219.91000366211)
('Accuracy:', 0.59287059)
('Lexicon length: ', 423)
('features length: ', 423)
Traceback (most recent call last):
  File "sentiment_demo.py", line 110, in <module>
train_neural_network(x)
  File "sentiment_demo.py", line 102, in train_neural_network
result = sess.run(tf.argmax(prediction.eval(feed_dict={x:features}), 1))
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 569, in eval
return _eval_using_default_session(self, feed_dict, self.graph, session)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 3741, in _eval_using_default_session
return session.run(tensors, feed_dict)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 778, in run
run_metadata_ptr)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 982, in _run
feed_dict_string, options, run_metadata)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1032, in _do_run
target_list, options, run_metadata)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1052, in _do_call
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: In[0] is not a matrix
 [[Node: MatMul = MatMul[T=DT_FLOAT, transpose_a=false, transpose_b=false, _device="/job:localhost/replica:0/task:0/gpu:0"](_recv_Placeholder_0/_23, Variable/read)]]
 [[Node: add/_25 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/cpu:0", send_device="/job:localhost/replica:0/task:0/gpu:0", send_device_incarnation=1, tensor_name="edge_4_add", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/cpu:0"]()]]

Caused by op u'MatMul', defined at:
  File "sentiment_demo.py", line 110, in <module>
    train_neural_network(x)
  File "sentiment_demo.py", line 58, in train_neural_network
    prediction = nueral_network_model(x)
  File "sentiment_demo.py", line 44, in nueral_network_model
    l1 = tf.add(tf.matmul(data, hidden_1_layer['weight']), hidden_1_layer['bias'])
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/ops/math_ops.py", line 1801, in matmul
a, b, transpose_a=transpose_a, transpose_b=transpose_b, name=name)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/ops/gen_math_ops.py", line 1263, in _mat_mul
transpose_b=transpose_b, name=name)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/op_def_library.py", line 768, in apply_op
op_def=op_def)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 2336, in create_op
original_op=self._default_original_op, op_def=op_def)
  File "/home/lsmpc/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1228, in __init__
self._traceback = _extract_stack()

InvalidArgumentError (see above for traceback): In[0] is not a matrix
 [[Node: MatMul = MatMul[T=DT_FLOAT, transpose_a=false, transpose_b=false, _device="/job:localhost/replica:0/task:0/gpu:0"](_recv_Placeholder_0/_23, Variable/read)]]
 [[Node: add/_25 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/cpu:0", send_device="/job:localhost/replica:0/task:0/gpu:0", send_device_incarnation=1, tensor_name="edge_4_add", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/cpu:0"]()]]

上面的错误具体指向这个:

Caused by op u'MatMul', defined at:
  File "sentiment_demo.py", line 110, in <module>
    train_neural_network(x)
  File "sentiment_demo.py", line 58, in train_neural_network
    prediction = nueral_network_model(x)
  File "sentiment_demo.py", line 44, in nueral_network_model
    l1 = tf.add(tf.matmul(data, hidden_1_layer['weight']), hidden_1_layer['bias'])

我是新手，无法修复它。

Answer 1

看起来您的特征形状不对。请试试这个:

    features = np.array(list(features)).reshape(1,-1)

您的模型接受批量数据，因此如果您只想运行一个预测，则需要将其重新整形为一批 1。祝你好运!