通过构建更复杂的深度学习模型可以提高分类的准确性，即分别基于TextCNN、TextRNN和TextRCNN三种算法实现中文文本分类。
项目地址：zz-zik/NLP-Application-and-Practice: 本项目将《自然语言处理与应用实战》原书中代码进行了实现，并在此基础上进行了改进。原书作者：韩少云、裴广战、吴飞等。 (github.com)
该项目目录如图：
实用工具
utils.py代码编写
# coding: UTF-8
import os
import torch
import numpy as np
import pickle as pkl
from tqdm import tqdm
import time
from datetime import timedelta
MAX_VOCAB_SIZE = 10000 # 词表长度限制
UNK, PAD = '<UNK>', '<PAD>' # 未知字，padding符号
def build_vocab(file_path, tokenizer, max_size, min_freq):
vocab_dic = {}
with open(file_path, 'r', encoding='UTF-8') as f:
for line in tqdm(f):
lin = line.strip()
if not lin:
continue
content = lin.split('\t')[0]
for word in tokenizer(content):
vocab_dic[word] = vocab_dic.get(word, 0) + 1
vocab_list = sorted([_ for _ in vocab_dic.items() if _[1] >= min_freq], key=lambda x: x[1],
reverse=True)[:max_size]
vocab_dic = {word_count[0]: idx for idx, word_count in enumerate(vocab_list)}
vocab_dic.update({UNK: len(vocab_dic), PAD: len(vocab_dic) + 1})
return vocab_dic
def build_dataset(config, ues_word):
if ues_word:
tokenizer = lambda x: x.split(' ') # 以空格隔开，word-level
else:
tokenizer = lambda x: [y for y in x] # char-level
if os.path.exists(config.vocab_path):
vocab = pkl.load(open(config.vocab_path, 'rb'))
else:
vocab = build_vocab(config.train_path, tokenizer=tokenizer, max_size=MAX_VOCAB_SIZE, min_freq=1)
pkl.dump(vocab, open(config.vocab_path, 'wb'))
print(f"Vocab size: {len(vocab)}")
def load_dataset(path, pad_size=32):
contents = []
with open(path, 'r', encoding='UTF-8') as f:
for line in tqdm(f):
lin = line.strip()
if not lin:
continue
content, label = lin.split('\t')
words_line = []
token = tokenizer(content)
seq_len = len(token)
if pad_size:
if len(token) < pad_size:
token.extend([PAD] * (pad_size - len(token)))
else:
token = token[:pad_size]
seq_len = pad_size
# word to id
for word in token:
words_line.append(vocab.get(word, vocab.get(UNK)))
contents.append((words_line, int(label), seq_len))
return contents # [([...], 0), ([...], 1), ...]
train = load_dataset(config.train_path, config.pad_size)
dev = load_dataset(config.dev_path, config.pad_size)
test = load_dataset(config.test_path, config.pad_size)
return vocab, train, dev, test, # predict
class DatasetIterater(object):
def __init__(self, batches, batch_size, device):
self.batch_size = batch_size
self.batches = batches
self.n_batches = len(batches) // batch_size
self.residue = False # 记录batch数量是否为整数
if len(batches) % self.n_batches != 0:
self.residue = True
self.index = 0
self.device = device
def _to_tensor(self, datas):
x = torch.LongTensor([_[0] for _ in datas]).to(self.device)
y = torch.LongTensor([_[1] for _ in datas]).to(self.device)
# pad前的长度(超过pad_size的设为pad_size)
seq_len = torch.LongTensor([_[2] for _ in datas]).to(self.device)
return (x, seq_len), y
def __next__(self):
if self.residue and self.index == self.n_batches:
batches = self.batches[self.index * self.batch_size: len(self.batches)]
self.index += 1
batches = self._to_tensor(batches)
return batches
elif self.index >= self.n_batches:
self.index = 0
raise StopIteration
else:
batches = self.batches[self.index * self.batch_size: (self.index + 1) * self.batch_size]
self.index += 1
batches = self._to_tensor(batches)
return batches
def __iter__(self):
return self
def __len__(self):
if self.residue:
return self.n_batches + 1
else:
return self.n_batches
def build_iterator(dataset, config, predict):
if predict==True:
config.batch_size = 1
iter = DatasetIterater(dataset, config.batch_size, config.device)
return iter
def get_time_dif(start_time):
"""获取已使用时间"""
end_time = time.time()
time_dif = end_time - start_time
return timedelta(seconds=int(round(time_dif)))
if __name__ == "__main__":
'''提取预训练词向量'''
# 下面的目录、文件名按需更改。
train_dir = "./THUCNews/data/train.txt"
vocab_dir = "./THUCNews/data/vocab.pkl"
pretrain_dir = "./THUCNews/data/sgns.sogou.char"
emb_dim = 300
filename_trimmed_dir = "./THUCNews/data/embedding_SougouNews"
if os.path.exists(vocab_dir):
word_to_id = pkl.load(open(vocab_dir, 'rb'))
else:
# tokenizer = lambda x: x.split(' ') # 以词为单位构建词表(数据集中词之间以空格隔开)
tokenizer = lambda x: [y for y in x] # 以字为单位构建词表
word_to_id = build_vocab(train_dir, tokenizer=tokenizer, max_size=MAX_VOCAB_SIZE, min_freq=1)
pkl.dump(word_to_id, open(vocab_dir, 'wb'))
embeddings = np.random.rand(len(word_to_id), emb_dim)
f = open(pretrain_dir, "r", encoding='UTF-8')
for i, line in enumerate(f.readlines()):
# if i == 0: # 若第一行是标题，则跳过
# continue
lin = line.strip().split(" ")
if lin[0] in word_to_id:
idx = word_to_id[lin[0]]
emb = [float(x) for x in lin[1:301]]
embeddings[idx] = np.asarray(emb, dtype='float32')
f.close()
np.savez_compressed(filename_trimmed_dir, embeddings=embeddings)
快速文本工具
utils_fasttext.py代码编写
# coding: UTF-8
import os
import torch
import numpy as np
import pickle as pkl
from tqdm import tqdm
import time
from datetime import timedelta
MAX_VOCAB_SIZE = 10000
UNK, PAD = '<UNK>', '<PAD>'
def build_vocab(file_path, tokenizer, max_size, min_freq):
vocab_dic = {}
with open(file_path, 'r', encoding='UTF-8') as f:
for line in tqdm(f):
lin = line.strip()
if not lin:
continue
content = lin.split('\t')[0]
for word in tokenizer(content):
vocab_dic[word] = vocab_dic.get(word, 0) + 1
vocab_list = sorted([_ for _ in vocab_dic.items() if _[1] >= min_freq], key=lambda x: x[1],
reverse=True)[:max_size]
vocab_dic = {word_count[0]: idx for idx, word_count in enumerate(vocab_list)}
vocab_dic.update({UNK: len(vocab_dic), PAD: len(vocab_dic) + 1})
return vocab_dic
def build_dataset(config, ues_word):
if ues_word:
tokenizer = lambda x: x.split(' ') # 以空格隔开，word-level
else:
tokenizer = lambda x: [y for y in x] # char-level
if os.path.exists(config.vocab_path):
vocab = pkl.load(open(config.vocab_path, 'rb'))
else:
vocab = build_vocab(config.train_path, tokenizer=tokenizer, max_size=MAX_VOCAB_SIZE, min_freq=1)
pkl.dump(vocab, open(config.vocab_path, 'wb'))
print(f"Vocab size: {len(vocab)}")
def biGramHash(sequence, t, buckets):
t1 = sequence[t - 1] if t - 1 >= 0 else 0
return (t1 * 14918087) % buckets
def triGramHash(sequence, t, buckets):
t1 = sequence[t - 1] if t - 1 >= 0 else 0
t2 = sequence[t - 2] if t - 2 >= 0 else 0
return (t2 * 14918087 * 18408749 + t1 * 14918087) % buckets
def load_dataset(path, pad_size=32):
contents = []
with open(path, 'r', encoding='UTF-8') as f:
for line in tqdm(f):
lin = line.strip()
if not lin:
continue
content, label = lin.split('\t')
words_line = []
token = tokenizer(content)
seq_len = len(token)
if pad_size:
if len(token) < pad_size:
token.extend([PAD] * (pad_size - len(token)))
else:
token = token[:pad_size]
seq_len = pad_size
# word to id
for word in token:
words_line.append(vocab.get(word, vocab.get(UNK)))
# fasttext ngram
buckets = config.n_gram_vocab
bigram = []
trigram = []
# ------ngram------
for i in range(pad_size):
bigram.append(biGramHash(words_line, i, buckets))
trigram.append(triGramHash(words_line, i, buckets))
# -----------------
contents.append((words_line, int(label), seq_len, bigram, trigram))
return contents # [([...], 0), ([...], 1), ...]
train = load_dataset(config.train_path, config.pad_size)
dev = load_dataset(config.dev_path, config.pad_size)
test = load_dataset(config.test_path, config.pad_size)
return vocab, train, dev, test
class DatasetIterater(object):
def __init__(self, batches, batch_size, device):
self.batch_size = batch_size
self.batches = batches
self.n_batches = len(batches) // batch_size
self.residue = False # 记录batch数量是否为整数
if len(batches) % self.n_batches != 0:
self.residue = True
self.index = 0
self.device = device
def _to_tensor(self, datas):
# xx = [xxx[2] for xxx in datas]
# indexx = np.argsort(xx)[::-1]
# datas = np.array(datas)[indexx]
x = torch.LongTensor([_[0] for _ in datas]).to(self.device)
y = torch.LongTensor([_[1] for _ in datas]).to(self.device)
bigram = torch.LongTensor([_[3] for _ in datas]).to(self.device)
trigram = torch.LongTensor([_[4] for _ in datas]).to(self.device)
# pad前的长度(超过pad_size的设为pad_size)
seq_len = torch.LongTensor([_[2] for _ in datas]).to(self.device)
return (x, seq_len, bigram, trigram), y
def __next__(self):
if self.residue and self.index == self.n_batches:
batches = self.batches[self.index * self.batch_size: len(self.batches)]
self.index += 1
batches = self._to_tensor(batches)
return batches
elif self.index >= self.n_batches:
self.index = 0
raise StopIteration
else:
batches = self.batches[self.index * self.batch_size: (self.index + 1) * self.batch_size]
self.index += 1
batches = self._to_tensor(batches)
return batches
def __iter__(self):
return self
def __len__(self):
if self.residue:
return self.n_batches + 1
else:
return self.n_batches
def build_iterator(dataset, config, predict):
if predict == True:
config.batch_size = 1
iter = DatasetIterater(dataset, config.batch_size, config.device)
return iter
def get_time_dif(start_time):
"""获取已使用时间"""
end_time = time.time()
time_dif = end_time - start_time
return timedelta(seconds=int(round(time_dif)))
if __name__ == "__main__":
'''提取预训练词向量'''
vocab_dir = "./THUCNews/data/vocab.pkl"
pretrain_dir = "./THUCNews/data/sgns.sogou.char"
emb_dim = 300
filename_trimmed_dir = "./THUCNews/data/vocab.embedding.sougou"
word_to_id = pkl.load(open(vocab_dir, 'rb'))
embeddings = np.random.rand(len(word_to_id), emb_dim)
f = open(pretrain_dir, "r", encoding='UTF-8')
for i, line in enumerate(f.readlines()):
# if i == 0: # 若第一行是标题，则跳过
# continue
lin = line.strip().split(" ")
if lin[0] in word_to_id:
idx = word_to_id[lin[0]]
emb = [float(x) for x in lin[1:301]]
embeddings[idx] = np.asarray(emb, dtype='float32')
f.close()
np.savez_compressed(filename_trimmed_dir, embeddings=embeddings)
数据训练
train_eval.py代码编写
# coding: UTF-8
import numpy as np
import torch
import torch.nn as nn
import torch.nn.functional as F
from sklearn import metrics
import time
from torch.utils.tensorboard import SummaryWriter
from utils import get_time_dif
# 权重初始化，默认xavier
def init_network(model, method='xavier', exclude='embedding', seed=123):
for name, w in model.named_parameters():
if exclude not in name:
if 'weight' in name:
if method == 'xavier':
nn.init.xavier_normal_(w)
elif method == 'kaiming':
nn.init.kaiming_normal_(w)
else:
nn.init.normal_(w)
elif 'bias' in name:
nn.init.constant_(w, 0)
else:
pass
def train(config, model, train_iter, dev_iter, test_iter):
start_time = time.time()
model.train()
optimizer = torch.optim.Adam(model.parameters(), lr=config.learning_rate)
# 学习率指数衰减，每次epoch：学习率 = gamma * 学习率
# scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.9)
total_batch = 0 # 记录进行到多少batch
dev_best_loss = float('inf')
last_improve = 0 # 记录上次验证集loss下降的batch数
flag = False # 记录是否很久没有效果提升
writer = SummaryWriter(log_dir=config.log_path + '/' + time.strftime('%m-%d_%H.%M', time.localtime()))
for epoch in range(config.num_epochs):
print('Epoch [{}/{}]'.format(epoch + 1, config.num_epochs))
# scheduler.step() # 学习率衰减
for i, (trains, labels) in enumerate(train_iter):
outputs = model(trains)
model.zero_grad()
loss = F.cross_entropy(outputs, labels)
loss.backward()
optimizer.step()
if total_batch % 100 == 0:
# 每多少轮输出在训练集和验证集上的效果
true = labels.data.cpu()
predic = torch.max(outputs.data, 1)[1].cpu()
train_acc = metrics.accuracy_score(true, predic)
dev_acc, dev_loss = evaluate(config, model, dev_iter)
if dev_loss < dev_best_loss:
dev_best_loss = dev_loss
torch.save(model.state_dict(), config.save_path)
improve = '*'
last_improve = total_batch
else:
improve = ''
time_dif = get_time_dif(start_time)
msg = 'Iter: {0:>6}, Train Loss: {1:>5.2}, Train Acc: {2:>6.2%}, Val Loss: {3:>5.2}, ' \
'Val Acc: {4:>6.2%}, Time: {5} {6}'
print(msg.format(total_batch, loss.item(), train_acc, dev_loss, dev_acc, time_dif, improve))
writer.add_scalar("loss/train", loss.item(), total_batch)
writer.add_scalar("loss/dev", dev_loss, total_batch)
writer.add_scalar("acc/train", train_acc, total_batch)
writer.add_scalar("acc/dev", dev_acc, total_batch)
model.train()
total_batch += 1
if total_batch - last_improve > config.require_improvement:
# 验证集loss超过1000batch没下降，结束训练
print("No optimization for a long time, auto-stopping...")
flag = True
break
if flag:
break
writer.close()
test(config, model, test_iter)
def test(config, model, test_iter):
# test
model.load_state_dict(torch.load(config.save_path))
model.eval()
start_time = time.time()
test_acc, test_loss, test_report, test_confusion = evaluate(config, model, test_iter, test=True)
msg = 'Test Loss: {0:>5.2}, Test Acc: {1:>6.2%}'
print(msg.format(test_loss, test_acc))
print("Precision, Recall and F1-Score...")
print(test_report)
print("Confusion Matrix...")
print(test_confusion)
time_dif = get_time_dif(start_time)
print("Time usage:", time_dif)
def evaluate(config, model, data_iter, test=False):
model.eval()
loss_total = 0
predict_all = np.array([], dtype=int)
labels_all = np.array([], dtype=int)
with torch.no_grad():
for texts, labels in data_iter:
outputs = model(texts)
loss = F.cross_entropy(outputs, labels)
loss_total += loss
labels = labels.data.cpu().numpy()
predic = torch.max(outputs.data, 1)[1].cpu().numpy()
labels_all = np.append(labels_all, labels)
predict_all = np.append(predict_all, predic)
acc = metrics.accuracy_score(labels_all, predict_all)
if test:
report = metrics.classification_report(labels_all, predict_all, target_names=config.class_list, digits=4)
confusion = metrics.confusion_matrix(labels_all, predict_all)
return acc, loss_total / len(data_iter), report, confusion
return acc, loss_total / len(data_iter)
数据测试
text_mixture_predict.py代码编写
# coding:utf-8
import torch
import numpy as np
import pickle as pkl
from importlib import import_module
from utils import build_iterator
import argparse
parser = argparse.ArgumentParser(description='Chinese Text Classification')
parser.add_argument('--model', type=str, required=True, help='choose a model: TextCNN, TextRNN, TextRCNN')
parser.add_argument('--embedding', default='pre_trained', type=str, help='random or pre_trained')
parser.add_argument('--word', default=False, type=bool, help='True for word, False for char')
args = parser.parse_args()
MAX_VOCAB_SIZE = 10000 # 词表长度限制
tokenizer = lambda x: [y for y in x] # char-level
UNK, PAD = '<UNK>', '<PAD>' # 未知字，padding符号
def load_dataset(content, vocab, pad_size=32):
contents = []
for line in content:
lin = line.strip()
if not lin:
continue
# content, label = lin.split('\t')
words_line = []
token = tokenizer(line)
seq_len = len(token)
if pad_size:
if len(token) < pad_size:
token.extend([PAD] * (pad_size - len(token)))
else:
token = token[:pad_size]
seq_len = pad_size
# word to id
for word in token:
words_line.append(vocab.get(word, vocab.get(UNK)))
contents.append((words_line, int(0), seq_len))
return contents # [([...], 0), ([...], 1), ...]
def match_label(pred, config):
label_list = config.class_list
return label_list[pred]
def final_predict(config, model, data_iter):
map_location = lambda storage, loc: storage
model.load_state_dict(torch.load(config.save_path, map_location=map_location))
model.eval()
predict_all = np.array([])
with torch.no_grad():
for texts, _ in data_iter:
outputs = model(texts)
pred = torch.max(outputs.data, 1)[1].cpu().numpy()
pred_label = [match_label(i, config) for i in pred]
predict_all = np.append(predict_all, pred_label)
return predict_all
def main(text):
dataset = 'THUCNews' # 数据集
# 搜狗新闻:embedding_SougouNews.npz, 腾讯:embedding_Tencent.npz, 随机初始化:random
embedding = 'embedding_SougouNews.npz'
if args.embedding == 'random':
embedding = 'random'
model_name = args.model # 'TextRCNN' # TextCNN, TextRNN, FastText, TextRCNN, TextRNN_Att, DPCNN, Transformer
x = import_module('models.' + model_name)
config = x.Config(dataset, embedding)
vocab = pkl.load(open(config.vocab_path, 'rb'))
content = load_dataset(text, vocab, 64)
predict = True
predict_iter = build_iterator(content, config, predict)
config.n_vocab = len(vocab)
model = x.Model(config).to(config.device)
result = final_predict(config, model, predict_iter)
for i, j in enumerate(result):
print('text:{}'.format(text[i]),'\t','label:{}'.format(j))
if __name__ == '__main__':
test = ['国考28日网上查报名序号查询后务必牢记报名参加2011年国家公务员的考生，如果您已通过资格审查，那么请于10月28日8：00后，登录考录专题网站查询自己的“关键数字”——报名序号。'
'国家公务员局等部门提醒：报名序号是报考人员报名确认和下载打印准考证等事项的重要依据和关键字，请务必牢记。此外，由于年龄在35周岁以上、40周岁以下的应届毕业硕士研究生和'
'博士研究生(非在职)，不通过网络进行报名，所以，这类人报名须直接与要报考的招录机关联系，通过电话传真或发送电子邮件等方式报名。',
'高品质低价格东芝L315双核本3999元作者：徐彬【北京行情】2月20日东芝SatelliteL300(参数图片文章评论)采用14.1英寸WXGA宽屏幕设计，配备了IntelPentiumDual-CoreT2390'
'双核处理器(1.86GHz主频/1MB二级缓存/533MHz前端总线)、IntelGM965芯片组、1GBDDR2内存、120GB硬盘、DVD刻录光驱和IntelGMAX3100集成显卡。目前，它的经销商报价为3999元。',
'国安少帅曾两度出山救危局他已托起京师一代才俊新浪体育讯随着联赛中的连续不胜，卫冕冠军北京国安的队员心里到了崩溃的边缘，俱乐部董事会连夜开会做出了更换主教练洪元硕的决定。'
'而接替洪元硕的，正是上赛季在李章洙下课风波中同样下课的国安俱乐部副总魏克兴。生于1963年的魏克兴球员时代并没有特别辉煌的履历，但也绝对称得上特别：15岁在北京青年队获青年'
'联赛最佳射手，22岁进入国家队，著名的5-19一战中，他是国家队的替补队员。',
'汤盈盈撞人心情未平复眼泛泪光拒谈悔意(附图)新浪娱乐讯汤盈盈日前醉驾撞车伤人被捕，原本要彩排《欢乐满东华2008》的她因而缺席，直至昨日(12月2日)，盈盈继续要与王君馨、马'
'赛、胡定欣等彩排，大批记者在电视城守候，她足足迟了约1小时才到场。全身黑衣打扮的盈盈，神情落寞、木无表情，回答记者问题时更眼泛泪光。盈盈因为迟到，向记者说声“不好意思”后'
'便急步入场，其助手坦言盈盈没什么可以讲。后来在《欢乐满东华2008》监制何小慧陪同下，盈盈接受简短访问，她小声地说：“多谢大家关心，交给警方处理了，不方便讲，',
'甲醇期货今日挂牌上市继上半年焦炭、铅期货上市后，酝酿已久的甲醇期货将在今日正式挂牌交易。基准价均为3050元／吨继上半年焦炭、铅期货上市后，酝酿已久的甲醇期货将在今日正式'
'挂牌交易。郑州商品交易所（郑商所）昨日公布首批甲醇期货8合约的上市挂牌基准价，均为3050元／吨。据此推算，买卖一手甲醇合约至少需要12200元。业内人士认为，作为国际市场上的'
'首个甲醇期货品种，其今日挂牌后可能会因炒新资金追捧而出现冲高走势，脉冲式行情过后可能有所回落，不过，投资者在上市初期应关注期现价差异常带来的无风险套利交易机会。']
main(test)
TextCNN
# coding: UTF-8
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
class Config(object):
"""配置参数"""
def __init__(self, dataset, embedding):
self.model_name = 'TextCNN'
self.train_path = dataset + '/data/train.txt' # 训练集
self.dev_path = dataset + '/data/dev.txt' # 验证集
self.test_path = dataset + '/data/test.txt' # 测试集
self.predict_path = dataset + '/data/predict.txt'
# self.class_list = [x.strip() for x in open(dataset + '/data/class.txt', encoding='utf-8').readlines()] # 类别名单
self.class_list = ['财经', '房产', '股票', '教育', '科技', '社会', '时政', '体育', '游戏','娱乐']
self.vocab_path = dataset + '/data/vocab.pkl' # 词表
self.save_path = dataset + '/saved_dict/' + self.model_name + '.ckpt' # 模型训练结果
self.log_path = dataset + '/log/' + self.model_name
self.embedding_pretrained = torch.tensor(
np.load(dataset + '/data/' + embedding)["embeddings"].astype('float32'))\
if embedding != 'random' else None # 预训练词向量
self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') # 设备
self.dropout = 0.5 # 随机失活
self.require_improvement = 1000 # 若超过1000batch效果还没提升，则提前结束训练
self.num_classes = len(self.class_list) # 类别数
self.n_vocab = 0 # 词表大小，在运行时赋值
self.num_epochs = 20 # epoch数
self.batch_size = 128 # mini-batch大小
self.pad_size = 32 # 每句话处理成的长度(短填长切)
self.learning_rate = 1e-3 # 学习率
self.embed = self.embedding_pretrained.size(1)\
if self.embedding_pretrained is not None else 300 # 字向量维度
self.filter_sizes = (2, 3, 4) # 卷积核尺寸
self.num_filters = 256 # 卷积核数量(channels数)
'''Convolutional Neural Networks for Sentence Classification'''
class Model(nn.Module):
def __init__(self, config):
super(Model, self).__init__()
if config.embedding_pretrained is not None:
self.embedding = nn.Embedding.from_pretrained(config.embedding_pretrained, freeze=False)
else:
self.embedding = nn.Embedding(config.n_vocab, config.embed, padding_idx=config.n_vocab - 1)
self.convs = nn.ModuleList(
[nn.Conv2d(1, config.num_filters, (k, config.embed)) for k in config.filter_sizes])
self.dropout = nn.Dropout(config.dropout)
self.fc = nn.Linear(config.num_filters * len(config.filter_sizes), config.num_classes)
def conv_and_pool(self, x, conv):
x = F.relu(conv(x)).squeeze(3)
x = F.max_pool1d(x, x.size(2)).squeeze(2)
return x
def forward(self, x):
out = self.embedding(x[0])
out = out.unsqueeze(1)
out = torch.cat([self.conv_and_pool(out, conv) for conv in self.convs], 1)
out = self.dropout(out)
out = self.fc(out)
return out
训练效果
| TextCNN训练效果 | ||||
| Test Time | 0:04:23 | |||
| Test Loaa: | 0.3 | Test Acc | 90.58% | |
| 类别 | Precision | recall | F1-score | support |
| 财经 | 0.9123 | 0.8950 | 0.9036 | 1000 |
| 房产 | 0.9043 | 0.9360 | 0.9199 | 1000 |
| 股票 | 0.8812 | 0.8230 | 0.8511 | 1000 |
| 教育 | 0.9540 | 0.9530 | 0.9535 | 1000 |
| 科技 | 0.8354 | 0.8880 | 0.9609 | 1000 |
| 社会 | 0.8743 | 0.9110 | 0.8923 | 1000 |
| 时政 | 0.8816 | 0.8860 | 0.8838 | 1000 |
| 体育 | 0.9682 | 0.9430 | 0.9554 | 1000 |
| 游戏 | 0.9249 | 0.9110 | 0.9179 | 1000 |
| 娱乐 | 0.9287 | 0.9120 | 0.9203 | 1000 |
| Accuracy | 0.9058 | 1000 | ||
| Macro avg | 0.9065 | 0.9058 | 0.9059 | 1000 |
| Weighted avg | 0.9065 | 0.9058 | 0.9059 | 1000 |
TextRNN
# coding: UTF-8
import torch
import torch.nn as nn
import numpy as np
class Config(object):
"""配置参数"""
def __init__(self, dataset, embedding):
self.model_name = 'TextRNN'
self.train_path = dataset + '/data/train.txt' # 训练集
self.dev_path = dataset + '/data/dev.txt' # 验证集
self.test_path = dataset + '/data/test.txt' # 测试集
# self.class_list = [x.strip() for x in open(dataset + '/data/class.txt', encoding='utf-8').readlines()]
self.class_list = ['体育', '军事', '娱乐', '政治', '教育', '灾难', '社会', '科技', '财经', '违法']
self.vocab_path = dataset + '/data/vocab.pkl' # 词表
self.save_path = dataset + '/saved_dict/' + self.model_name + '.ckpt' # 模型训练结果
self.log_path = dataset + '/log/' + self.model_name
self.embedding_pretrained = torch.tensor(
np.load(dataset + '/data/' + embedding)["embeddings"].astype('float32'))\
if embedding != 'random' else None # 预训练词向量
self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') # 设备
self.dropout = 0.5 # 随机失活
self.require_improvement = 1000 # 若超过1000batch效果还没提升，则提前结束训练
self.num_classes = len(self.class_list) # 类别数
self.n_vocab = 0 # 词表大小，在运行时赋值
self.num_epochs = 20 # epoch数
self.batch_size = 128 # mini-batch大小
self.pad_size = 32 # 每句话处理成的长度(短填长切)
self.learning_rate = 1e-3 # 学习率
self.embed = self.embedding_pretrained.size(1)\
if self.embedding_pretrained is not None else 300 # 字向量维度, 若使用了预训练词向量，则维度统一
self.hidden_size = 128 # lstm隐藏层
self.num_layers = 2 # lstm层数
'''Recurrent Neural Network for Text Classification with Multi-Task Learning'''
class Model(nn.Module):
def __init__(self, config):
super(Model, self).__init__()
if config.embedding_pretrained is not None:
self.embedding = nn.Embedding.from_pretrained(config.embedding_pretrained, freeze=False)
else:
self.embedding = nn.Embedding(config.n_vocab, config.embed, padding_idx=config.n_vocab - 1)
self.lstm = nn.LSTM(config.embed, config.hidden_size, config.num_layers,
bidirectional=True, batch_first=True, dropout=config.dropout)
self.fc = nn.Linear(config.hidden_size * 2, config.num_classes)
def forward(self, x):
x, _ = x
out = self.embedding(x) # [batch_size, seq_len, embeding]=[128, 32, 300]
out, _ = self.lstm(out)
out = self.fc(out[:, -1, :]) # 句子最后时刻的 hidden state
return out
'''变长RNN，效果差不多，甚至还低了点...'''
# def forward(self, x):
# x, seq_len = x
# out = self.embedding(x)
# _, idx_sort = torch.sort(seq_len, dim=0, descending=True) # 长度从长到短排序（index）
# _, idx_unsort = torch.sort(idx_sort) # 排序后，原序列的 index
# out = torch.index_select(out, 0, idx_sort)
# seq_len = list(seq_len[idx_sort])
# out = nn.utils.rnn.pack_padded_sequence(out, seq_len, batch_first=True)
# # [batche_size, seq_len, num_directions * hidden_size]
# out, (hn, _) = self.lstm(out)
# out = torch.cat((hn[2], hn[3]), -1)
# # out, _ = nn.utils.rnn.pad_packed_sequence(out, batch_first=True)
# out = out.index_select(0, idx_unsort)
# out = self.fc(out)
# return out
训练效果
| TextRNN训练效果 | ||||
| Test Time | 0:05:07 | |||
| Test Loaa: | 0.29 | Test Acc | 91.03% | |
| 类别 | Precision | recall | F1-score | support |
| 财经 | 0.9195 | 0.8790 | 0.8988 | 1000 |
| 房产 | 0.9181 | 0.9190 | 0.9185 | 1000 |
| 股票 | 0.8591 | 0.8290 | 0.8438 | 1000 |
| 教育 | 0.9349 | 0.9480 | 0.9414 | 1000 |
| 科技 | 0.8642 | 0.8720 | 0.8681 | 1000 |
| 社会 | 0.9190 | 0.9080 | 0.9135 | 1000 |
| 时政 | 0.8578 | 0.8990 | 0.8779 | 1000 |
| 体育 | 0.9690 | 0.9690 | 0.9690 | 1000 |
| 游戏 | 0.9454 | 0.9350 | 0.9402 | 1000 |
| 娱乐 | 0.9175 | 0.9450 | 0.9310 | 1000 |
| Accuracy | 0.9103 | 1000 | ||
| Macro avg | 0.9104 | 0.9103 | 0.9102 | 1000 |
| Weighted avg | 0.9104 | 0.9103 | 0.9102 | 1000 |
TextRNN网络的训练效果最好，准确率达到了91.03%，明显高于TextCNN网络的效果。
TextRCNN
# coding: UTF-8
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
class Config(object):
"""配置参数"""
def __init__(self, dataset, embedding):
self.model_name = 'TextRCNN'
self.train_path = dataset + '/data/train.txt' # 训练集
self.dev_path = dataset + '/data/dev.txt' # 验证集
self.test_path = dataset + '/data/test.txt' # 测试集
self.class_list = [x.strip() for x in open(
dataset + '/data/class.txt', encoding='utf-8').readlines()] # 类别名单
self.vocab_path = dataset + '/data/vocab.pkl' # 词表
self.save_path = dataset + '/saved_dict/' + self.model_name + '.ckpt' # 模型训练结果
self.log_path = dataset + '/log/' + self.model_name
self.embedding_pretrained = torch.tensor(
np.load(dataset + '/data/' + embedding)["embeddings"].astype('float32'))\
if embedding != 'random' else None # 预训练词向量
self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') # 设备
self.dropout = 1.02 # 随机失活
self.require_improvement = 1000 # 若超过1000batch效果还没提升，则提前结束训练
self.num_classes = len(self.class_list) # 类别数
self.n_vocab = 0 # 词表大小，在运行时赋值
self.num_epochs = 20 # epoch数
self.batch_size = 128 # mini-batch大小
self.pad_size = 32 # 每句话处理成的长度(短填长切)
self.learning_rate = 1e-3 # 学习率
self.embed = self.embedding_pretrained.size(1)\
if self.embedding_pretrained is not None else 300 # 字向量维度, 若使用了预训练词向量，则维度统一
self.hidden_size = 256 # lstm隐藏层
self.num_layers = 1 # lstm层数
'''Recurrent Convolutional Neural Networks for Text Classification'''
class Model(nn.Module):
def __init__(self, config):
super(Model, self).__init__()
if config.embedding_pretrained is not None:
self.embedding = nn.Embedding.from_pretrained(config.embedding_pretrained, freeze=False)
else:
self.embedding = nn.Embedding(config.n_vocab, config.embed, padding_idx=config.n_vocab - 1)
self.lstm = nn.LSTM(config.embed, config.hidden_size, config.num_layers,
bidirectional=True, batch_first=True, dropout=config.dropout)
self.maxpool = nn.MaxPool1d(config.pad_size)
self.fc = nn.Linear(config.hidden_size * 2 + config.embed, config.num_classes)
def forward(self, x):
x, _ = x
embed = self.embedding(x) # [batch_size, seq_len, embeding]=[64, 32, 64]
out, _ = self.lstm(embed)
out = torch.cat((embed, out), 2)
out = F.relu(out)
out = out.permute(0, 2, 1)
out = self.maxpool(out).squeeze()
out = self.fc(out)
return out
训练效果
| TextRCNN训练效果 | ||||
| Test Time | 0:03:20 | |||
| Test Loaa: | 0.29 | Test Acc | 90.96% | |
| 类别 | Precision | recall | F1-score | support |
| 财经 | 0.9134 | 0.8970 | 0.9051 | 1000 |
| 房产 | 0.9051 | 0.9350 | 0.9198 | 1000 |
| 股票 | 0.8658 | 0.8320 | 0.8485 | 1000 |
| 教育 | 0.9295 | 0.9500 | 0.9397 | 1000 |
| 科技 | 0.8352 | 0.8770 | 0.8556 | 1000 |
| 社会 | 0.8993 | 0.9290 | 0.9139 | 1000 |
| 时政 | 0.8921 | 0.9680 | 0.8799 | 1000 |
| 体育 | 0.9851 | 0.9670 | 0.9743 | 1000 |
| 游戏 | 0.9551 | 0.9140 | 0.9341 | 1000 |
| 娱乐 | 0.9233 | 0.9270 | 0.9251 | 1000 |
| Accuracy | 0.9096 | 1000 | ||
| Macro avg | 0.9101 | 0.9096 | 0.9096 | 1000 |
| Weighted avg | 0.9101 | 0.9096 | 0.9096 | 1000 |
TextRCNN网络的效果为90.96%，与TextCNN网络模型效果相近。
| mode | time | Cpu |
| TextCNN | 0:04:23 | CORE i5 |
| TextRNN | 0:05:07 | |
| TextRCNN | 0:03:20 |
通过分别对模型TextCNN、TextRNN、TextRCNN和不同的硬件环境进行实验，分别对实验结果中的训练时间、准确率、召回率和F1值进行比较，进一步确定哪个模型在给定数据集和硬件环境下表现最佳。最终发现TextRNN的训练结果最好，但所需的时间也是最久的，而TextRCNN模型的训练结果与TextCNN几乎相同，但TextRCNN所需的时间最少。
