第N4周：使用Word2vec实现文本分类

news2024/9/20 6:31:23

结合Word2Vec文本内容（第1列）预测文本标签（第2列）
尝试根据第2周的内容独立实现，尽可能的不看本文的代码
进一步了解并学习Word2Vec
任务说明：
本次将加入Word2vec使用PyTorch实现中文文本分类，Word2Vec则是其中的一种词嵌入方法，是一种用于生成词向量的浅层神经网络模型，由Tomas Mikolov及其团队于2013年提出。**Word2Vec通过学习大量文本数据，将每个单词表示为一个连续的向量，这些向量可以捕捉单词之间的语义和句法关系。**更详细的内容可见训练营内的NLP基础知识，数据示例如下：
在这里插入图片描述

二、数据预处理

1.加载数据

warnings.filterwarnings("ignore")  #忽略警告信息
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print('device =', device)


print('STEP.1', '-' * 19)
''' 加载自定义中文数据 '''
train_data = pd.read_csv('./data/train.csv', sep='\t', header=None)
print(train_data.head())
# 构造数据集迭代器
def coustom_data_iter(texts, labels):
    for x, y in zip(texts, labels):
        yield x, y
x = train_data[0].values[:]
# 多类标签的one-hot展开
y = train_data[1].values[:]

2.构建词典

print('STEP.2', '-' * 19)
''' 构建词典 '''
# 训练Word2Vec浅层神经网络模型
w2v = Word2Vec(vector_size=100,  #是指特征向量的维度，默认为100。
               min_count=3)      #可以对字典做截断. 词频少于min_count次数的单词会被丢弃掉, 默认值为5。
w2v.build_vocab(x)
w2v.train(x,
          total_examples=w2v.corpus_count,
          epochs=20)

# 将文本转化为向量
def average_vec(text):
    vec = np.zeros((1,100))
    for word in text:
        try:
            vec += w2v.wv[word].reshape((1,100))
        except KeyError:
            continue
    return vec

# 将词向量保存为ndarray
x_vec = np.concatenate([average_vec(z) for z in x])
print('len(x) =', len(x), 'len(x_vec) =', len(x_vec))
# 保存Word2Vec模型及词向量
w2v.save('output/w2v_model.pkl')
train_iter = coustom_data_iter(x_vec, y)

3.生成数据批次和迭代器

print('STEP.3', '-' * 19)
''' 准备数据处理管道 '''
label_name = list(set(train_data[1].values[:]))
print(label_name)
text_pipeline  = lambda x: average_vec(x)
label_pipeline = lambda x: label_name.index(x)
print('你在干嘛', text_pipeline('你在干嘛'))
print('Travel-Query', label_pipeline('Travel-Query'))
''' 生成数据批次和迭代器 '''
def collate_batch(batch):
    label_list, text_list = [], []
    #offsets = [0]
    for (_text, _label) in batch:
        # 标签列表
        label_list.append(label_pipeline(_label))
        # 文本列表
        processed_text = torch.tensor(text_pipeline(_text), dtype=torch.float32)
        text_list.append(processed_text)
        #offsets.append(processed_text.size(0))
    
    label_list = torch.tensor(label_list, dtype=torch.int64)  # torch.Size([64])
    text_list = torch.cat(text_list)  # 若干tensor组成的列表变成一个tensor
    #offsets = torch.tensor(offsets[:-1]).cumsum(dim=0)  # torch.Size([64])
    return text_list.to(device), label_list.to(device)  # , offsets.to(device)
# 数据加载器
#dataloader = DataLoader(train_iter, batch_size=8, shuffle=False, collate_fn=collate_batch)

二、模型构建

1.搭建模型

print('STEP.4', '-' * 19)
''' 搭建文本分类模型 '''
class TextClassificationModel(nn.Module):
    def __init__(self, num_class):
        super(TextClassificationModel, self).__init__()
        self.fc = nn.Linear(100, num_class)
    
    def forward(self, text):
        output = self.fc(text)
        return output

2.初始化模型

print('STEP.5 Initialize', '-' * 19)
''' 初始化实例 '''
num_class  = len(label_name)
vocab_size = 100000  # 词典大小
emsize     = 12      # 嵌入的维度
model      = TextClassificationModel(num_class).to(device)

3.定义训练与评估函数

''' 训练函数 '''
def train(dataloader):
    model.train()  # 训练模式
    total_acc, train_loss, total_count = 0, 0, 0
    log_interval = 50
    start_time = time.time()
    
    for idx, (text, label) in enumerate(dataloader):
        optimizer.zero_grad()  # grad属性归零
        predited_label = model(text)
        loss = criterion(predited_label, label)
        loss.backward()  # 反向传播
        torch.nn.utils.clip_grad_norm_(model.parameters(), 0.1)  # 梯度裁剪
        optimizer.step()  # 每一步自动更新
        # 记录acc与loss
        total_acc += (predited_label.argmax(1) == label).sum().item()
        train_loss  += loss.item()
        total_count += label.size(0)
        if idx % log_interval == 0 and idx > 0:
            elapsed = time.time() - start_time
            print('| epoch {:3d} | {:5d}/{:5d} batches, train_acc {:8.3f} train_loss {:8.3f}'.format(epoch, idx, len(dataloader), total_acc/total_count, train_loss/total_count))
            total_acc, train_loss, total_count = 0, 0, 0
            start_time = time.time()


''' 评估函数 '''
def evaluate(dataloader):
    model.eval()  # 切换为测试模式
    total_acc, train_loss, total_count = 0, 0, 0
    
    with torch.no_grad():
        for idx, (text, label) in enumerate(dataloader):
            predited_label = model(text)
            loss = criterion(predited_label, label)  # 计算loss值
            # 记录测试数据
            total_acc += (predited_label.argmax(1) == label).sum().item()
            train_loss  += loss.item()
            total_count += label.size(0)
    return total_acc/total_count, train_loss/total_count

三、训练模型

1.拆分数据集并运行模型

''' 开始训练 '''
if __name__ == '__main__':
    # 超参数(Hyperparameters)
    EPOCHS     = 10  # epoch
    LR         = 5   # learning rate
    BATCH_SIZE = 64  # batch size for training
    
    criterion = torch.nn.CrossEntropyLoss()
    optimizer = torch.optim.SGD(model.parameters(), lr=LR)
    scheduler = torch.optim.lr_scheduler.StepLR(optimizer, 1.0, gamma=0.1)
    total_accu = None
    # 构建数据集
    train_iter = coustom_data_iter(train_data[0].values[:], train_data[1].values[:])
    train_dataset = list(train_iter)
    # 划分数据集
    num_train = int(len(train_dataset) * 0.8)
    split_train_, split_valid_ = random_split(train_dataset, [num_train, len(train_dataset) - num_train])
    # 加载数据集
    train_dataloader = DataLoader(split_train_, batch_size=BATCH_SIZE, shuffle=True, collate_fn=collate_batch)      # shuffle表示随机打乱
    valid_dataloader = DataLoader(split_valid_, batch_size=BATCH_SIZE, shuffle=True, collate_fn=collate_batch)
    
    for epoch in range(1, EPOCHS + 1):
        epoch_start_time = time.time()
        train(train_dataloader)
        accu_val, loss_val = evaluate(valid_dataloader)
        # 获取当前的学习率
        lr = optimizer.state_dict()['param_groups'][0]['lr']
        if total_accu is not None and total_accu > accu_val:
            scheduler.step()
        else:
            total_accu = accu_val
        print('-' * 59)
        print('| end of epoch {:3d} | time: {:5.2f}s | '
              'valid_acc {:8.3f} valid_loss {:8.3f} | lr {:8.6f}'.format(epoch, time.time()-epoch_start_time, accu_val, loss_val, lr))
        print('-' * 59)
    
    torch.save(model.state_dict(), 'output\\model_TextClassification.pth')
   
    print('Checking the results of test dataset.')
    accu_test, loss_test = evaluate(valid_dataloader)
    print('test accuracy {:8.3f}, test loss {:8.3f}'.format(accu_test, loss_test))

2.测试指定数据

''' 预测函数 '''
def predict(text, text_pipeline):
    with torch.no_grad():
        text = torch.tensor(text_pipeline(text), dtype=torch.float32)
        print(text.shape)
        output = model(text)
        return output.argmax(1).item()


''' 以下是预测 '''
if __name__=='__main__':
    model.load_state_dict(torch.load('output\\model_TextClassification.pth'))
    #label_name = ['Alarm-Update', 'Other', 'Audio-Play', 'Calendar-Query', 'HomeAppliance-Control', 'Radio-Listen', 'Travel-Query', 'Video-Play', 'TVProgram-Play', 'FilmTele-Play', 'Weather-Query', 'Music-Play']
    
    ex_text_str = "随便播放一首专辑阁楼里的佛里的歌"
    #ex_text_str = "还有双鸭山到淮阴的汽车票吗13号的"
    model = model.to("cpu")
    
    print("该文本的类别是：%s" % label_name[predict(ex_text_str, text_pipeline)])