-
Notifications
You must be signed in to change notification settings - Fork 14
/
SA_predict.py
42 lines (42 loc) · 1.73 KB
/
SA_predict.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
# coding: utf-8
import os
import tensorflow as tf
import numpy as np
import tensorflow.contrib.keras as kr
from model.cnn_model import TCNNConfig, TextCNN
from model.data_processing import read_category, read_vocab
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
base_dir = 'data'
vocab_dir = os.path.join(base_dir, 'vocab.txt')
save_dir = 'checkpoints/textcnn'
save_path = os.path.join(save_dir, 'best_validation')
class CnnModel:
def __init__(self):
self.config = TCNNConfig()
self.categories, self.cat_to_id = read_category()
self.words, self.word_to_id = read_vocab(vocab_dir)
self.config.vocab_size = len(self.words)
self.model = TextCNN(self.config)
self.session = tf.Session()
self.session.run(tf.global_variables_initializer())
# 加载模型
saver = tf.train.Saver()
saver.restore(sess=self.session, save_path=save_path)
def emotion_score(self, message):
data = [self.word_to_id[x] for x in message if x in self.word_to_id]
feed_dict = {
self.model.input_x: kr.preprocessing.sequence.pad_sequences([data], self.config.seq_length),
self.model.keep_prob: 1.0
}
# 类别概率的输出
predictions = self.session.run(self.model.softmax_tensor1, feed_dict=feed_dict)
return np.squeeze(predictions)[1]
# if __name__ == '__main__':
# import time
# for i in open('data/demo_predict.txt', 'r', encoding='utf-8'):
# st = time.clock()
# tf.reset_default_graph()
# test_text = i.strip()
# predict_score = CnnModel().emotion_score(test_text)
# print('{},情感分析结果:{}'.format(test_text, predict_score))
# print('time used:{}'.format(time.clock()-st))