def __init__(self,
embedding_dim=100,
batch_size=64,
n_hidden=100,
learning_rate=0.01,
n_class=3,
max_sentence_len=50,
l2_reg=0.,
display_step=4,
n_iter=100,
type_=''):
self.embedding_dim = embedding_dim
self.batch_size = batch_size
self.n_hidden = n_hidden
self.learning_rate = learning_rate
self.n_class = n_class
self.max_sentence_len = max_sentence_len
self.l2_reg = l2_reg
self.display_step = display_step
self.n_iter = n_iter
self.type_ = type_
self.word_id_mapping, self.w2v = load_word_embedding(
FLAGS.word_id_file_path, FLAGS.embedding_file_path,
self.embedding_dim)
# self.word_embedding = tf.constant(self.w2v, dtype=tf.float32, name='word_embedding')
self.word_embedding = tf.Variable(self.w2v,
dtype=tf.float32,
name='word_embedding')
# self.word_id_mapping = load_word_id_mapping(FLAGS.word_id_file_path)
# self.word_embedding = tf.Variable(
# tf.random_uniform([len(self.word_id_mapping), self.embedding_dim], -0.1, 0.1), name='word_embedding')
self.aspect_id_mapping, self.aspect_embed = load_aspect2id(
FLAGS.aspect_id_file_path, self.word_id_mapping, self.w2v,
self.embedding_dim)
self.aspect_embedding = tf.Variable(self.aspect_embed,
dtype=tf.float32,
name='aspect_embedding')
self.keep_prob1 = tf.placeholder(tf.float32)
self.keep_prob2 = tf.placeholder(tf.float32)
with tf.name_scope('inputs'):
self.x = tf.placeholder(tf.int32, [None, self.max_sentence_len],
name='x')
self.y = tf.placeholder(tf.int32, [None, self.n_class], name='y')
self.sen_len = tf.placeholder(tf.int32, None, name='sen_len')
self.aspect_id = tf.placeholder(tf.int32, None, name='aspect_id')
with tf.name_scope('weights'):
self.weights = {
'softmax':
tf.get_variable(
name='softmax_w',
shape=[self.n_hidden, self.n_class],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
}
with tf.name_scope('biases'):
self.biases = {
'softmax':
tf.get_variable(
name='softmax_b',
shape=[self.n_class],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
}
self.W = tf.get_variable(
name='W',
shape=[
self.n_hidden + self.embedding_dim,
self.n_hidden + self.embedding_dim
],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.w = tf.get_variable(
name='w',
shape=[self.n_hidden + self.embedding_dim, 1],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.Wp = tf.get_variable(
name='Wp',
shape=[self.n_hidden, self.n_hidden],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.Wx = tf.get_variable(
name='Wx',
shape=[self.n_hidden, self.n_hidden],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
def __init__(self,
embedding_dim=100,
batch_size=64,
n_hidden=100,
learning_rate=0.01,
n_class=3,
max_sentence_len=50,
l2_reg=0.,
display_step=4,
n_iter=100,
type_=''):
self.embedding_dim = embedding_dim #300
self.batch_size = batch_size #25
self.n_hidden = n_hidden #300
self.learning_rate = learning_rate #0.01
self.n_class = n_class #3
self.max_sentence_len = max_sentence_len #80
self.l2_reg = l2_reg #0.001
self.display_step = display_step #4
self.n_iter = n_iter #20
self.type_ = type_ #AT
self.word_id_mapping, self.w2v = load_word_embedding(
FLAGS.word_id_file_path, FLAGS.embedding_file_path,
self.embedding_dim)
# dict(3909) 3910 * 300 word->id 路径 词嵌入的路径 词嵌入维度：300
# self.word_embedding = tf.constant(self.w2v, dtype=tf.float32, name='word_embedding')
self.word_embedding = tf.Variable(
self.w2v, dtype=tf.float32,
name='word_embedding') # 定义word_embedding变量
# self.word_id_mapping = load_word_id_mapping(FLAGS.word_id_file_path)
# self.word_embedding = tf.Variable(
# tf.random_uniform([len(self.word_id_mapping), self.embedding_dim], -0.1, 0.1), name='word_embedding')
self.aspect_id_mapping, self.aspect_embed = load_aspect2id(
FLAGS.aspect_id_file_path, self.word_id_mapping, self.w2v,
self.embedding_dim)
# dict(1219) 1220 * 300
self.aspect_embedding = tf.Variable(
self.aspect_embed, dtype=tf.float32,
name='aspect_embedding') #定义word_embedding变量
#定义droupout占位符
self.keep_prob1 = tf.placeholder(tf.float32, name="dropout_keep_prob1")
self.keep_prob2 = tf.placeholder(tf.float32, name="dropout_keep_prob2")
with tf.name_scope('inputs'):
self.x = tf.placeholder(tf.int32, [None, self.max_sentence_len],
name='x') #25 * 80
#print (self.max_sentence_len) #80
#print ('sxl================')
self.y = tf.placeholder(tf.int32, [None, self.n_class],
name='y') #25 * 3
self.sen_len = tf.placeholder(tf.int32, None,
name='sen_len') #list(25)
self.aspect_id = tf.placeholder(tf.int32, None,
name='aspect_id') #list(25)
with tf.name_scope('weights'):
self.weights = {
'softmax':
tf.get_variable(
name='softmax_w',
shape=[self.n_hidden, self.n_class], #300 * 3
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
}
with tf.name_scope('biases'):
self.biases = {
'softmax':
tf.get_variable(
name='softmax_b',
shape=[self.n_class], # 3
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
}
self.W = tf.get_variable(
name='W',
shape=[
self.n_hidden + self.embedding_dim,
self.n_hidden + self.embedding_dim
], #600 * 600
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.w = tf.get_variable(
name='w',
shape=[self.n_hidden + self.embedding_dim, 1], #600 * 1
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.Wp = tf.get_variable(
name='Wp',
shape=[self.n_hidden, self.n_hidden], #300 * 300
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
self.Wx = tf.get_variable(
name='Wx',
shape=[self.n_hidden, self.n_hidden], #300 * 300
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg))
def __init__(self, config):
self.embedding_dim = config.embedding_dim
self.batch_size = config.batch_size
self.n_hidden = config.n_hidden
self.learning_rate = config.learning_rate
self.n_class = config.n_class
self.max_len = config.max_len
self.l2_reg = config.l2_reg
self.display_step = config.display_step
self.n_iter = config.n_iter
self.embedding_file = config.embedding_file_path
self.word2id_file = config.word_id_file_path
self.aspect_id_file = config.aspect_id_file_path
self.train_file = config.train_file_path
self.test_file = config.test_file_path
self.val_file = config.validate_file_path
self.word2id, self.w2v = load_w2v(self.embedding_file,
self.embedding_dim)
self.word_embedding = tf.constant(self.w2v, name='word_embedding')
self.aspect2id, self.a2v = load_aspect2id(self.aspect_id_file,
self.word2id, self.w2v,
self.embedding_dim)
self.aspect_embedding = tf.constant(self.a2v, name='aspect_embedding')
with tf.name_scope('inputs'):
self.x = tf.placeholder(tf.int32, [None, self.max_len], name='x')
self.y = tf.placeholder(tf.float32, [None, self.n_class], name='y')
self.sen_len = tf.placeholder(tf.int32, None, name='sen_len')
self.aspect_id = tf.placeholder(tf.int32, None, name='aspect_id')
self.position = tf.placeholder(tf.int32, [None, self.max_len],
name='position')
with tf.name_scope('GRU'):
self.w_r = tf.get_variable(
name='W_r',
shape=[2 * self.n_hidden + 1, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
self.u_r = tf.get_variable(
name='U_r',
shape=[self.n_hidden, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
self.w_z = tf.get_variable(
name='W_z',
shape=[2 * self.n_hidden + 1, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
self.u_z = tf.get_variable(
name='U_z',
shape=[self.n_hidden, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
self.w_x = tf.get_variable(
name='W_x',
shape=[self.n_hidden, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
self.w_g = tf.get_variable(
name='W_g',
shape=[2 * self.n_hidden + 1, self.n_hidden],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer(uniform=True))
import tensorflow as tf
import numpy as np
from utils import load_w2v, batch_index, load_word_embedding, load_aspect2id, load_inputs_twitter_at
x_raw = ["$T$ is always fresh and hot - ready to eat !", "food"]
y_test = [1]
word_id_mapping, w2v = load_word_embedding(
'data/restaurant/word_id_new.txt',
'data/restaurant/rest_2014_word_embedding_300_new.txt', 300)
# dict(3909) 3910 * 300
aspect_id_mapping, aspect_embed = load_aspect2id(
'data/restaurant/aspect_id_new.txt', word_id_mapping, w2v, 300)
# dict(1219) 1220 * 300
# print (aspect_id_mapping['food'])
# print ('sxlllllllllll')
def change_y_to_onehot(y):
class_set = set([1, -1, 0])
n_class = 3
y_onehot_mapping = {0: 0, 1: 1, -1: 2}
#print (y_onehot_mapping)
onehot = []
for label in y:
tmp = [0] * n_class
tmp[y_onehot_mapping[label]] = 1
onehot.append(tmp)
return np.asarray(onehot, dtype=np.int32)
def __init__(self, embedding_dim=100, batch_size=64, n_hidden=100, learning_rate=0.01,
n_class=3, max_sentence_len=50, l2_reg=0., display_step=4, n_iter=100, type_=''):
self.embedding_dim = embedding_dim
self.batch_size = batch_size
self.n_hidden = n_hidden
self.learning_rate = learning_rate
self.n_class = n_class
self.max_sentence_len = max_sentence_len
self.l2_reg = l2_reg
self.display_step = display_step
self.n_iter = n_iter
self.type_ = type_
self.word_id_mapping, self.w2v = load_word_embedding(FLAGS.word_id_file_path, FLAGS.embedding_file_path, self.embedding_dim)
# self.word_embedding = tf.constant(self.w2v, dtype=tf.float32, name='word_embedding')
self.word_embedding = tf.Variable(self.w2v, dtype=tf.float32, name='word_embedding')
# self.word_id_mapping = load_word_id_mapping(FLAGS.word_id_file_path)
# self.word_embedding = tf.Variable(
# tf.random_uniform([len(self.word_id_mapping), self.embedding_dim], -0.1, 0.1), name='word_embedding')
self.aspect_id_mapping, self.aspect_embed = load_aspect2id(FLAGS.aspect_id_file_path, self.word_id_mapping, self.w2v, self.embedding_dim)
self.aspect_embedding = tf.Variable(self.aspect_embed, dtype=tf.float32, name='aspect_embedding')
self.keep_prob1 = tf.placeholder(tf.float32)
self.keep_prob2 = tf.placeholder(tf.float32)
with tf.name_scope('inputs'):
self.x = tf.placeholder(tf.int32, [None, self.max_sentence_len], name='x')
self.y = tf.placeholder(tf.int32, [None, self.n_class], name='y')
self.sen_len = tf.placeholder(tf.int32, None, name='sen_len')
self.aspect_id = tf.placeholder(tf.int32, None, name='aspect_id')
with tf.name_scope('weights'):
self.weights = {
'softmax': tf.get_variable(
name='softmax_w',
shape=[self.n_hidden, self.n_class],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)
}
with tf.name_scope('biases'):
self.biases = {
'softmax': tf.get_variable(
name='softmax_b',
shape=[self.n_class],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)
}
self.W = tf.get_variable(
name='W',
shape=[self.n_hidden + self.embedding_dim, self.n_hidden + self.embedding_dim],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)
self.w = tf.get_variable(
name='w',
shape=[self.n_hidden + self.embedding_dim, 1],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)
self.Wp = tf.get_variable(
name='Wp',
shape=[self.n_hidden, self.n_hidden],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)
self.Wx = tf.get_variable(
name='Wx',
shape=[self.n_hidden, self.n_hidden],
initializer=tf.random_uniform_initializer(-0.01, 0.01),
regularizer=tf.contrib.layers.l2_regularizer(self.l2_reg)
)