本文深入探讨了TextCNN在文本分类任务中的应用，包括模型原理、网络结构定义、数据预处理及训练流程。通过实际代码示例，展示了如何在TensorFlow中实现TextCNN，适用于情感分析、主题分类等场景。
TextCNN
论文地址：Convolutional Neural Networks for Sentence Classification
论文解读： A Sensitivity Analysis of (and Practitioners’ Guide to) Convolutional Neural Networks for Sentence Classification 这是一篇使用TextCNN时进行调参参考很好的一篇文章，具体我在另外一篇博客中进行了简要总结。
论文开源实现：
- Convolutional Neural Network for Text Classification in Tensorflow
- Implementing a CNN for Text Classification in TensorFlow
- 基于TensorFlow在中文数据集上的简化实现
- Python (2.7) and requires Theano (0.7) and Word2vec
- Keras实现
其他参考： - 用tf的VocabularyProcessor创建词汇表vocab
- tflearn的VocabularyProcessor用法：建立中文词汇表和把文本转为词ID序列
- https://blog.youkuaiyun.com/The_lastest/article/details/81771723
- Understanding Convolutional Neural Networks for NLP
一.CNN用于NLP原理概述
模型如下：
CNN最先用于计算机视觉领域。大多数NLP任务的输入不是图像像素，而是以矩阵形式表示的句子或文档。矩阵的每一行对应一个标记，通常是一个单词，但也可以是一个字符。也就是说，每一行都是表示一个单词的向量。通常，这些向量是单词嵌入(低维表示)，如word2vec或GloVe，但它们也可以是一个热向量，将单词编入词汇表。对于使用100维嵌入的10个单词的句子，将得到一个10*100矩阵作为输入，这种结构类似于图像。在CV中，filter在图像的局部块上滑动，但是在NLP中，我们通常使用filter在矩阵(单词)的整行上滑动。因此，filter的“宽度”通常与输入矩阵的宽度相同。高度或区域大小可能不同，但通常一次滑动窗口超过2-5个单词。综上所述，一个用于NLP的卷积神经网络可能是这样的(图片来源于A Sensitivity Analysis of (and Practitioners’ Guide to) Convolutional Neural Networks for Sentence Classification ):
对计算机视觉，位置不变性和局部合成对图像有直观的意义，但对自然语言处理则不然。你可能很在意一个单词在句子中的出现位置。相邻的像素可能在语义上是相关的(同一对象的一部分)，但对于单词并不总是如此。在许多语言中，短语的某些部分可以用其他几个词隔开，组成方面也不明显。很明显，单词在某些方面是有组成的，比如形容词修饰名词，但是这到底是如何工作的呢?更高层次的表示实际上是什么意思并不像在计算机视觉的例子中那样明显。
CNNs的一个重要论点是它们速度很快。卷积是计算机图形学的核心部分，在gpu的硬件中有针对其的实现。与n-gram相比，CNNs在表示方面也很有效。由于词汇量很大，计算任何超过3-gram的计算代价都会变得很高。即使谷歌也不能提供超过5-gram。卷积过滤器自动学习良好的表示，而不需要表示整个词汇表。尺寸大于5的过滤器是完全合理的。我个人认为第一层中的许多过滤器捕获的特性与n-g非常相似(但不受限制)，但以更紧凑的方式表示出来。
最适合CNNs的似乎是分类任务，比如情绪分析、垃圾邮件检测或主题分类。卷积和池操作会丢失关于单词的本地顺序的信息，因此序列标记(如PoS标记或实体提取)更难适应纯CNN体系结构(尽管不是不可能，如可以向输入添加位置特性)。
Tensorflow实现
本代码是对原始论文的简单复现，数据集使用原始论文中的Movie Review data from Rotten Tomatoes，该数据集包含10662个示例复习句子，一半是肯定句，一半是否定句。数据集的词汇表大小约为20k。注意，由于这个数据集非常小，很可能会使模型过拟合。此外，数据集没有附带官方的训练/测试分割，所以只使用10%的数据作为开发集。不使用预训练的word2vec向量来进行单词嵌入。相反，这里从头开始学习嵌入。不会对权重向量强制L2范数约束，根据对卷积神经网络进行句子分类的敏感性分析发现，这些约束对最终结果几乎没有影响。原文采用静态和非静态两种输入数据通道进行实验，这里只用一个通道。代码已经进行了详细的注释，不再过多解释：
#data_helper.py:数据预处理
import numpy as np
import re
def clean_str(string):
"""
Tokenization/string cleaning for all datasets except for SST.
Original taken from https://github.com/yoonkim/CNN_sentence/blob/master/process_data.py
re.sub(pattern, repl, string, count=0, flags=0) ：
pattern：匹配的模式
repl:replacement，替换后的字符串
string:要被处理，要被替换的那个string字符串
count:替换个数。默认为0，表示每个匹配项都替换
"""
#除A-Za-z0-9(),!?'`外的字符，去除
string = re.sub(r"[^A-Za-z0-9(),!?\'\`]", " ", string)
#匹配 '跟一个空格字符，在'前加个空格
string = re.sub(r"\'s", " \'s", string)
# 've 在'前加个空格
string = re.sub(r"\'ve", " \'ve", string)
#n'跟制表符，在n前加空格
string = re.sub(r"n\'t", " n\'t", string)
# 're 在'前加个空格
string = re.sub(r"\'re", " \'re", string)
#'d在'前加个空格
string = re.sub(r"\'d", " \'d", string)
#'ll在'前加个空格
string = re.sub(r"\'ll", " \'ll", string)
# ,前后各加空格
string = re.sub(r",", " , ", string)
#!前后 各加空格
string = re.sub(r"!", " ! ", string)
# (前后加空格
string = re.sub(r"\(", " \( ", string)
#)前后加空格
string = re.sub(r"\)", " \) ", string)
#?前后 加空格
string = re.sub(r"\?", " \? ", string)
#两个以上连续的空白符，删除
string = re.sub(r"\s{2,}", " ", string)
return string.strip().lower()
def load_data_and_labels(positive_data_file, negative_data_file):
"""
Loads MR polarity data from files, splits the data into words and generates labels.
Returns split sentences and labels.
"""
# Load data from files
positive_examples = list(open(positive_data_file, "r", encoding='utf-8').readlines())
positive_examples = [s.strip() for s in positive_examples]
negative_examples = list(open(negative_data_file, "r", encoding='utf-8').readlines())
negative_examples = [s.strip() for s in negative_examples]
# Split by words
x_text = positive_examples + negative_examples
x_text = [clean_str(sent) for sent in x_text]
# Generate labels
positive_labels = [[0, 1] for _ in positive_examples]
negative_labels = [[1, 0] for _ in negative_examples]
y = np.concatenate([positive_labels, negative_labels], 0)
return [x_text, y]
def batch_iter(data, batch_size, num_epochs, shuffle=True):
"""
Generates a batch iterator for a dataset.
"""
data = np.array(data)
data_size = len(data)
num_batches_per_epoch = int((len(data)-1)/batch_size) + 1
for epoch in range(num_epochs):
# Shuffle the data at each epoch
if shuffle:
shuffle_indices = np.random.permutation(np.arange(data_size))
shuffled_data = data[shuffle_indices]
else:
shuffled_data = data
for batch_num in range(num_batches_per_epoch):
start_index = batch_num * batch_size
end_index = min((batch_num + 1) * batch_size, data_size)
yield shuffled_data[start_index:end_index]
#text_cnn.py：网络结构定义
import tensorflow as tf
import numpy as np
class TextCNN(object):
"""
A CNN for text classification.
Uses an embedding layer, followed by a convolutional, max-pooling and softmax layer.
sequence_length：通过pad将句子长度统一到59
num_classes：输出的分类数
"""
def __init__(
self, sequence_length, num_classes, vocab_size,
embedding_size, filter_sizes, num_filters, l2_reg_lambda=0.0):
# Placeholders for input, output and dropout
self.input_x = tf.placeholder(tf.int32, [None, sequence_length], name="input_x")
self.input_y = tf.placeholder(tf.float32, [None, num_classes], name="input_y")
self.dropout_keep_prob = tf.placeholder(tf.float32, name="dropout_keep_prob")
# Keeping track of l2 regularization loss (optional)
l2_loss = tf.constant(0.0)
# Embedding layer
#embedding操作没有实现GPU支持，因此使用CPU
with tf.device('/cpu:0'), tf.name_scope("embedding"):
self.W = tf.Variable(
tf.random_uniform([vocab_size, embedding_size], -1.0, 1.0),
name="W")
# 输出为[Batch_size, sequence_length, embedding_size]
self.embedded_chars = tf.nn.embedding_lookup(self.W, self.input_x)
#Tensorflow的conv2d操作需要四维数据：[batch, in_height, in_width, in_channels]
#对应的添加一个维度，构成[batch,width,sequence_length,embedding_size，1]
self.embedded_chars_expanded = tf.expand_dims(self.embedded_chars, -1)
# Create a convolution + maxpool layer for each filter size
pooled_outputs = []
#使用不容region_size(height)大小的卷积核
for i, filter_size in enumerate(filter_sizes):
with tf.name_scope("conv-maxpool-%s" % filter_size):
# Convolution Layer
filter_shape = [filter_size, embedding_size, 1, num_filters]
W = tf.Variable(tf.truncated_normal(filter_shape, stddev=0.1), name="W")
b = tf.Variable(tf.constant(0.1, shape=[num_filters]), name="b")
#输入参数：X:[batch, in_height, in_width, in_channels]
# filter_shape:[filter_height, filter_width, in_channels, out_channels]
# strides:[1, height, width, 1]
# padding:"SAME"或"VALID"
#每个filter滑过整个embeding(width)，只是在滑过的词语数量(height)上不同。
#应用窄卷积后输出结果维度为[1, sequence_length - filter_size + 1, 1, 1]
conv = tf.nn.conv2d(
self.embedded_chars_expanded,
W,
strides=[1, 1, 1, 1],
padding="VALID", #without padding the edges
name="conv")
# Apply nonlinearity
h = tf.nn.relu(tf.nn.bias_add(conv, b), name="relu")
# Maxpooling over the outputs
#池化后输出维度 [batch_size, 1, 1, num_filters]
pooled = tf.nn.max_pool(
h,
ksize=[1, sequence_length - filter_size + 1, 1, 1],
strides=[1, 1, 1, 1],
padding='VALID',
name="pool")
pooled_outputs.append(pooled)#[batch_size, 1, 1, num_filters]
# Combine all the pooled features
num_filters_total = num_filters * len(filter_sizes)
self.h_pool = tf.concat(pooled_outputs, 3)#在最后一个维度联接
self.h_pool_flat = tf.reshape(self.h_pool, [-1, num_filters_total])
# Add dropout
with tf.name_scope("dropout"):
self.h_drop = tf.nn.dropout(self.h_pool_flat, self.dropout_keep_prob)
# Final (unnormalized) scores and predictions
with tf.name_scope("output"):
W = tf.get_variable(
"W",
shape=[num_filters_total, num_classes],
initializer=tf.contrib.layers.xavier_initializer())
b = tf.Variable(tf.constant(0.1, shape=[num_classes]), name="b")
l2_loss += tf.nn.l2_loss(W)
l2_loss += tf.nn.l2_loss(b)
self.scores = tf.nn.xw_plus_b(self.h_drop, W, b, name="scores")
self.predictions = tf.argmax(self.scores, 1, name="predictions")
# Calculate mean cross-entropy loss
with tf.name_scope("loss"):
losses = tf.nn.softmax_cross_entropy_with_logits(logits=self.scores, labels=self.input_y)
self.loss = tf.reduce_mean(losses) + l2_reg_lambda * l2_loss
# Accuracy
with tf.name_scope("accuracy"):
correct_predictions = tf.equal(self.predictions, tf.argmax(self.input_y, 1))
self.accuracy = tf.reduce_mean(tf.cast(correct_predictions, "float"), name="accuracy")
#! /usr/bin/env python
#train.py 网络训练
import tensorflow as tf
import numpy as np
import os
import time
import datetime
import data_helpers
from text_cnn import TextCNN
from tensorflow.contrib import learn
# Parameters
# ==================================================
# Data loading params
tf.flags.DEFINE_float("dev_sample_percentage", .1, "Percentage of the training data to use for validation")
tf.flags.DEFINE_string("positive_data_file", "./data/rt-polaritydata/rt-polarity.pos", "Data source for the positive data.")
tf.flags.DEFINE_string("negative_data_file", "./data/rt-polaritydata/rt-polarity.neg", "Data source for the negative data.")
# Model Hyperparameters
tf.flags.DEFINE_integer("embedding_dim", 128, "Dimensionality of character embedding (default: 128)")
tf.flags.DEFINE_string("filter_sizes", "3,4,5", "Comma-separated filter sizes (default: '3,4,5')")
tf.flags.DEFINE_integer("num_filters", 128, "Number of filters per filter size (default: 128)")
tf.flags.DEFINE_float("dropout_keep_prob", 0.5, "Dropout keep probability (default: 0.5)")
tf.flags.DEFINE_float("l2_reg_lambda", 0.0, "L2 regularization lambda (default: 0.0)")
# Training parameters
tf.flags.DEFINE_integer("batch_size", 64, "Batch Size (default: 64)")
tf.flags.DEFINE_integer("num_epochs", 200, "Number of training epochs (default: 200)")
tf.flags.DEFINE_integer("evaluate_every", 100, "Evaluate model on dev set after this many steps (default: 100)")
tf.flags.DEFINE_integer("checkpoint_every", 100, "Save model after this many steps (default: 100)")
tf.flags.DEFINE_integer("num_checkpoints", 5, "Number of checkpoints to store (default: 5)")
# Misc Parameters
tf.flags.DEFINE_boolean("allow_soft_placement", True, "Allow device soft device placement")
tf.flags.DEFINE_boolean("log_device_placement", False, "Log placement of ops on devices")
FLAGS = tf.flags.FLAGS
# FLAGS._parse_flags()
# print("\nParameters:")
# for attr, value in sorted(FLAGS.__flags.items()):
# print("{}={}".format(attr.upper(), value))
# print("")
def preprocess():
# Data Preparation
# ==================================================
# Load data
print("Loading data...")
x_text, y = data_helpers.load_data_and_labels(FLAGS.positive_data_file, FLAGS.negative_data_file)
# Build vocabulary
max_document_length = max([len(x.split(" ")) for x in x_text])
#VocabularyProcessor（max_document_length,min_frequency=0,vocabulary=None, tokenizer_fn=None)
#将文本转换为词语index列表
vocab_processor = learn.preprocessing.VocabularyProcessor(max_document_length)
x = np.array(list(vocab_processor.fit_transform(x_text)))
# Randomly shuffle data
np.random.seed(10)
shuffle_indices = np.random.permutation(np.arange(len(y)))
x_shuffled = x[shuffle_indices]
y_shuffled = y[shuffle_indices]
# Split train/test set
# TODO: This is very crude, should use cross-validation
#原文中使用的是cross-validation
dev_sample_index = -1 * int(FLAGS.dev_sample_percentage * float(len(y)))
x_train, x_dev = x_shuffled[:dev_sample_index], x_shuffled[dev_sample_index:]
y_train, y_dev = y_shuffled[:dev_sample_index], y_shuffled[dev_sample_index:]
del x, y, x_shuffled, y_shuffled
print("Vocabulary Size: {:d}".format(len(vocab_processor.vocabulary_)))
print("Train/Dev split: {:d}/{:d}".format(len(y_train), len(y_dev)))
return x_train, y_train, vocab_processor, x_dev, y_dev
def train(x_train, y_train, vocab_processor, x_dev, y_dev):
# Training
# ==================================================
with tf.Graph().as_default():
#命令行运行程序的参数
session_conf = tf.ConfigProto(
allow_soft_placement=FLAGS.allow_soft_placement, #让Tensorflow自动选择最合适的device进行运算
log_device_placement=FLAGS.log_device_placement) #指定Tensorflow使用GPU或CPU运算
sess = tf.Session(config=session_conf)
with sess.as_default():
cnn = TextCNN(
sequence_length=x_train.shape[1],
num_classes=y_train.shape[1],
vocab_size=len(vocab_processor.vocabulary_),
embedding_size=FLAGS.embedding_dim,
filter_sizes=list(map(int, FLAGS.filter_sizes.split(","))),
num_filters=FLAGS.num_filters,
l2_reg_lambda=FLAGS.l2_reg_lambda)
# Define Training procedure
global_step = tf.Variable(0, name="global_step", trainable=False)
optimizer = tf.train.AdamOptimizer(1e-3)
#variable总是有值的，但gradient可能是None
#计算loss中可训练的var_list中的梯度。 相当于minimize()的第一步，返回(gradient, variable)对的list。
grads_and_vars = optimizer.compute_gradients(cnn.loss)
#minimize()的第二部分，返回一个执行梯度更新的操作。
train_op = optimizer.apply_gradients(grads_and_vars, global_step=global_step)
# Keep track of gradient values and sparsity (optional)
grad_summaries = []
for g, v in grads_and_vars:
if g is not None:
#将输入的一个任意大小和形状的张量压缩成一个由宽度和数量组成的直方图数据结构
grad_hist_summary = tf.summary.histogram("{}/grad/hist".format(v.name), g)
#tf.nn.zero_fraction()的作用是将输入的Tensor中0元素在所有元素中所占的比例计算并返回
sparsity_summary = tf.summary.scalar("{}/grad/sparsity".format(v.name), tf.nn.zero_fraction(g))
grad_summaries.append(grad_hist_summary)
grad_summaries.append(sparsity_summary)
grad_summaries_merged = tf.summary.merge(grad_summaries)
# Output directory for models and summaries
timestamp = str(int(time.time()))
out_dir = os.path.abspath(os.path.join(os.path.curdir, "runs", timestamp))
print("Writing to {}\n".format(out_dir))
# Summaries for loss and accuracy
loss_summary = tf.summary.scalar("loss", cnn.loss)
acc_summary = tf.summary.scalar("accuracy", cnn.accuracy)
# Train Summaries
train_summary_op = tf.summary.merge([loss_summary, acc_summary, grad_summaries_merged])
train_summary_dir = os.path.join(out_dir, "summaries", "train")
train_summary_writer = tf.summary.FileWriter(train_summary_dir, sess.graph)
# Dev summaries
dev_summary_op = tf.summary.merge([loss_summary, acc_summary])
dev_summary_dir = os.path.join(out_dir, "summaries", "dev")
dev_summary_writer = tf.summary.FileWriter(dev_summary_dir, sess.graph)
# Checkpoint directory. Tensorflow assumes this directory already exists so we need to create it
checkpoint_dir = os.path.abspath(os.path.join(out_dir, "checkpoints"))
checkpoint_prefix = os.path.join(checkpoint_dir, "model")
if not os.path.exists(checkpoint_dir):
os.makedirs(checkpoint_dir)
saver = tf.train.Saver(tf.global_variables(), max_to_keep=FLAGS.num_checkpoints)
# Write vocabulary
vocab_processor.save(os.path.join(out_dir, "vocab"))
# Initialize all variables
sess.run(tf.global_variables_initializer())
def train_step(x_batch, y_batch):
"""
A single training step
"""
feed_dict = {
cnn.input_x: x_batch,
cnn.input_y: y_batch,
cnn.dropout_keep_prob: FLAGS.dropout_keep_prob
}
_, step, summaries, loss, accuracy = sess.run(
[train_op, global_step, train_summary_op, cnn.loss, cnn.accuracy],
feed_dict)
time_str = datetime.datetime.now().isoformat()#返回格式如'YYYY-MM-DD’的字符串；
print("{}: step {}, loss {:g}, acc {:g}".format(time_str, step, loss, accuracy))
train_summary_writer.add_summary(summaries, step)
def dev_step(x_batch, y_batch, writer=None):
"""
Evaluates model on a dev set
"""
feed_dict = {
cnn.input_x: x_batch,
cnn.input_y: y_batch,
cnn.dropout_keep_prob: 1.0
}
step, summaries, loss, accuracy = sess.run(
[global_step, dev_summary_op, cnn.loss, cnn.accuracy],
feed_dict)
time_str = datetime.datetime.now().isoformat()
print("{}: step {}, loss {:g}, acc {:g}".format(time_str, step, loss, accuracy))
if writer:
writer.add_summary(summaries, step)
# Generate batches
batches = data_helpers.batch_iter(
list(zip(x_train, y_train)), FLAGS.batch_size, FLAGS.num_epochs)
# Training loop. For each batch...
for batch in batches:
x_batch, y_batch = zip(*batch)
train_step(x_batch, y_batch)
current_step = tf.train.global_step(sess, global_step)
if current_step % FLAGS.evaluate_every == 0:
print("\nEvaluation:")
dev_step(x_dev, y_dev, writer=dev_summary_writer)
print("")
if current_step % FLAGS.checkpoint_every == 0:
path = saver.save(sess, checkpoint_prefix, global_step=current_step)
print("Saved model checkpoint to {}\n".format(path))
def main(argv=None):
x_train, y_train, vocab_processor, x_dev, y_dev = preprocess()
train(x_train, y_train, vocab_processor, x_dev, y_dev)
if __name__ == '__main__':
tf.app.run()
扫一扫
810
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
