51dev.com IT技术开发者社区

51dev.com 技术开发者社区

指针生成网络(Pointer-Generator-Network)原理与实战

代码星球阅读(1890)2020-04-03 收藏0次评论

指针生成网络(Pointer-Generator-Network)原理与实战

 

阅读目录

  • 0 前言
  • 1 Baseline sequence-to-sequence
  • 2 Pointer-Generator-Network
  • 3  Coverage mechanism
  • 4 实战部分
  • 4.1 DataSet
  • 4.2 Experiments
  • 4.3 Evaluation
  • 4.4 Results
  • 5 References
回到顶部

0 前言

      本文主要内容:介绍Pointer-Generator-Network在文本摘要任务中的背景模型架构与原理在中英文数据集上实战效果与评估,最后得出结论。参考的《Get To The Point: Summarization with Pointer-Generator Networks》以及多篇博客均在文末给出连接,文中使用数据集已上传百度网盘代码已传至GitHub,读者可以在文中找到相应连接,实际操作过程中确实遇到很多坑,并未在文中一一指明,有兴趣的读者可以留言一起交流。由于水平有限,请读者多多指正。

  随着互联网飞速发展,产生了越来越多的文本数据,文本信息过载问题日益严重,对各类文本进行一个“降 维”处理显得非常必要,文本摘要便是其中一个重要的手段。文本摘要旨在将文本或文本集合转换为包含关键信息的简短摘要。按照输出类型可分为抽取式摘要和生成式摘要。抽取式摘要从源文档中抽取关键句和关键词组成摘要,摘要全部来源于原文生成式摘要根据原文,允许生成新的词语、原文本中没有的短语来组成摘要。

  指针生成网络属于生成式模型。

  仅用Neural sequence-to-sequence模型可以实现生成式摘要,但存在两个问题:

    1. 可能不准确地再现细节, 无法处理词汇不足(OOV)单词;

    2. 倾向于重复自己

    原文是(they are liable to reproducefactual details inaccurately, and they tendto repeat themselves.)

  指针生成网络(Pointer-Generator-Network)从两个方面进行了改进

    1. 该网络通过指向(pointer)从源文本中复制单词,有助于准确地复制信息,同时保留通过生成器产生新单词的能力;

    2. 使用coverage机制来跟踪已总结的内容,防止重复。 

  接下来从下面几个部分介绍Pointer-Generator-Network原理:

    1. Baseline sequence-to-sequence;

    2. Pointer-Generator-Network;

    3. Coverage Mechanism。

回到顶部

1 Baseline sequence-to-sequence

  Pointer-Generator Networks是在Baseline sequence-to-sequence模型的基础上构建的,我们首先Baseline seq2seq+attention。其架构图如下:

 

  该模型可以关注原文本中的相关单词以生成新单词进行概括。比如:模型可能注意到原文中的"victorious"" win"这个两个单词,在摘要"Germany beat Argentina 2-0"中生成了新的单词beat 。

  Seq2Seq的模型结构是经典的Encoder-Decoder模型,即先用Encoder将原文本编码成一个中间层的隐藏状态,然后用Decoder来将该隐藏状态解码成为另一个文本。Baseline Seq2Seq在Encoder端是一个双向的LSTM,这个双向的LSTM可以捕捉原文本的长距离依赖关系以及位置信息,编码时词嵌入经过双向LSTM后得到编码状态 hihi 。在Decoder端解码器是一个单向的LSTM,训练阶段时参考摘要词依次输入(测试阶段时是上一步的生成词),在时间步 tt得到解码状态 stst 。使用hihi和stst得到该时间步原文第 ii个词注意力权重。

eti=vTtanh(Whhi+Wsst+battn)eit=vTtanh(Whhi+Wsst+battn) at=softmax(et)at=softmax(et)

  得到的注意力权重和 hihi加权求和得到重要的上下文向量 ht(contextvector)ht∗(contextvector):

 

ht=iatihiht∗=∑iaithi

 

  htht∗可以看成是该时间步通读了原文的固定尺寸的表征。然后将 stst和 htht∗ 经过两层线性层得到单词表分布 PvocabPvocab:

 

Pvocab=softmax(V(V[st,ht]+b)+b)Pvocab=softmax(V′(V[st,ht∗]+b)+b′)

 

  其中 [st,ht][st,ht∗]是拼接。这样再通过sofmaxsofmax得到了一个概率分布,就可以预测需要生成的词:

 

P(w)=Pvocab(w)P(w)=Pvocab(w)

 

  在训练阶段,时间步 tt 时的损失为: 

 

losst=logP(wt)losst=−logP(wt∗)

 

  那么原输入序列的整体损失为: 

 

loss=1Tt=0Tlosstloss=1T∑t=0Tlosst

 

 

回到顶部

2 Pointer-Generator-Network

  原文中的Pointer-Generator Networks是一个混合了 Baseline seq2seq和PointerNetwork的网络,它具有Baseline seq2seq的生成能力和PointerNetwork的Copy能力。该网络的结构如下:

如何权衡一个词应该是生成的还是复制的?

  原文中引入了一个权重 pgenpgen 。

  从Baseline seq2seq的模型结构中得到了stst 和htht∗,和解码器输入 xtxt 一起来计算 pgenpgen :

 

pgen=σ(wThht+wTsst+wTxxt+bptr)pgen=σ(wh∗Tht∗+wsTst+wxTxt+bptr)

 

  这时,会扩充单词表形成一个更大的单词表--扩充单词表(将原文当中的单词也加入到其中),该时间步的预测词概率为:

 

P(w)=pgenPvocab(w)+(1pgen)i:wi=watiP(w)=pgenPvocab(w)+(1−pgen)∑i:wi=wait

 

  其中 atiait 表示的是原文档中的词。我们可以看到解码器一个词的输出概率有其是否拷贝是否生成的概率和决定。当一个词不出现在常规的单词表上时 Pvocab(w)Pvocab(w) 为0,当该词不出现在文档中i:wi=wati∑i:wi=wait为0。

回到顶部

3  Coverage mechanism

  原文的特色是运用了Coverage Mechanism来解决重复生成文本的问题,下图反映了前两个模型与添加了Coverage Mechanism生成摘要的结果:

  蓝色的字体表示的是参考摘要,三个模型的生成摘要的结果差别挺大;

  红色字体表明了不准确的摘要细节生成(UNK未登录词,无法解决OOV问题);

  绿色的字体表明了模型生成了重复文本。

  为了解决此问题--Repitition,原文使用了在机器翻译中解决“过翻译”和“漏翻译”的机制--Coverage Mechanism

  具体实现上,就是将先前时间步的注意力权重加到一起得到所谓的覆盖向量 ct(coveragevector)ct(coveragevector),用先前的注意力权重决策来影响当前注意力权重的决策,这样就避免在同一位置重复,从而避免重复生成文本。计算上,先计算coverage vector ctct:

 

ct=t=0t1atct=∑t′=0t−1at′

 

  然后添加到注意力权重的计算过程中,ctct用来计算 etieit:

 

eti=vTtanh(Whhi+Wsst+wccti+battn)eit=vTtanh(Whhi+Wsst+wccit+battn)

 

  同时,为coverage vector添加损失是必要的,coverage loss计算方式为:

 

covlosst=imin(ati,cti)covlosst=∑imin(ait,cit)

 

  这样coverage loss是一个有界的量  covlosstiati=1covlosst≤∑iait=1 。因此最终的LOSS为:

 

losst=logP(wt)+λimin(ati,cti)losst=−logP(wt∗)+λ∑imin(ait,cit)  

 

回到顶部

4 实战部分

回到顶部

4.1 DataSet

英文数据集: cnn dailymail数据集,地址:https://github.com/becxer/cnn-dailymail/。

中文数据集:新浪微博摘要数据集,这是中文数据集,有679898条文本及摘要。

中英文数据集均可从这里下载,链接:https://pan.baidu.com/s/18ykewFUrTLzW8R84bF42pg  密码:9yqt。

回到顶部

4.2 Experiments

  试验环境:centos7.4/python3.6/tensorflow1.12.0  GPU:Tesla-K40m-12G*4   代码参考:python3 tensorflow版本。调试时候各种报错,所以需要debug。

  改动后的代码已上传至GitHub:https://github.com/zingp/NLP/tree/master/P007PytorchPointerGeneratorNetwork。

  中文数据集预处理代码:

  第一部分是对原始数据进行分词,划分训练集测试集,并保存文件。

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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 import os import sys import time import jieba   ARTICLE_FILE = "./data/weibo_news/train_text.txt" SUMMARRY_FILE = "./data/weibo_news/train_label.txt"   TRAIN_FILE = "./data/weibo_news/train_art_summ_prep.txt" VAL_FILE = "./data/weibo_news/val_art_summ_prep.txt"   def timer(func):     def wrapper(*args, **kwargs):         start = time.time()         = func(*args, **kwargs)         end = time.time()         cost = end - start         print(f"Cost time: {cost} s")         return r     return wrapper   @timer def load_data(filename):     """加载数据文件,对文本进行分词"""     data_list = []     with open(filename, 'r', encoding= 'utf-8') as f:         for line in f:             # jieba.enable_parallel()             words = jieba.cut(line.strip())             word_list = list(words)             # jieba.disable_parallel()             data_list.append(' '.join(word_list).strip())     return data_list   def build_train_val(article_data, summary_data, train_num=600_000):     """划分训练和验证数据"""     train_list = []     val_list = []     = 0     for text, summ in zip(article_data, summary_data):         += 1         if n <= train_num:             train_list.append(text)             train_list.append(summ)         else:             val_list.append(text)             val_list.append(summ)     return train_list, val_list   def save_file(filename, li):     """预处理后的数据保存到文件"""     with open(filename, 'w+', encoding='utf-8') as f:         for item in li:             f.write(item + ' ')     print(f"Save {filename} ok.")   if __name__ == '__main__':     article_data = load_data(ARTICLE_FILE)     # 大概耗时10分钟     summary_data = load_data(SUMMARRY_FILE)     TRAIN_SPLIT = 600_000     train_list, val_list = build_train_val(article_data, summary_data, train_num=TRAIN_SPLIT)     save_file(TRAIN_FILE, train_list)     save_file(VAL_FILE, val_list) 

第二部分是将文件打包,生成模型能够加载的二进制文件。

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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 import os import struct import collections from tensorflow.core.example import example_pb2   # 经过分词处理后的训练数据与测试数据文件 TRAIN_FILE = "./data/weibo_news/train_art_summ_prep.txt" VAL_FILE = "./data/weibo_news/val_art_summ_prep.txt"   # 文本起始与结束标志 SENTENCE_START = '<s>' SENTENCE_END = '</s>'   VOCAB_SIZE = 50_000  # 词汇表大小 CHUNK_SIZE = 1000    # 每个分块example的数量,用于分块的数据   # tf模型数据文件存放目录 FINISHED_FILE_DIR = './data/weibo_news/finished_files' CHUNKS_DIR = os.path.join(FINISHED_FILE_DIR, 'chunked')   def chunk_file(finished_files_dir, chunks_dir, name, chunk_size):     """构建二进制文件"""     in_file = os.path.join(finished_files_dir, '%s.bin' % name)     print(in_file)     reader = open(in_file, "rb")     chunk = 0     finished = False     while not finished:         chunk_fname = os.path.join(chunks_dir, '%s_%03d.bin' % (name, chunk))  # 新的分块         with open(chunk_fname, 'wb') as writer:             for in range(chunk_size):                 len_bytes = reader.read(8)                 if not len_bytes:                     finished = True                     break                 str_len = struct.unpack('q', len_bytes)[0]                 example_str = struct.unpack('%ds' % str_len, reader.read(str_len))[0]                 writer.write(struct.pack('q', str_len))                 writer.write(struct.pack('%ds' % str_len, example_str))             chunk += 1     def chunk_all():     # 创建一个文件夹来保存分块     if not os.path.isdir(CHUNKS_DIR):         os.mkdir(CHUNKS_DIR)     # 将数据分块     for name in ['train''val']:         print("Splitting %s data into chunks..." % name)         chunk_file(FINISHED_FILE_DIR, CHUNKS_DIR, name, CHUNK_SIZE)     print("Saved chunked data in %s" % CHUNKS_DIR)     def read_text_file(text_file):     """从预处理好的文件中加载数据"""     lines = []     with open(text_file, "r", encoding='utf-8') as f:         for line in f:             lines.append(line.strip())     return lines     def write_to_bin(input_file, out_file, makevocab=False):     """生成模型需要的文件"""     if makevocab:         vocab_counter = collections.Counter()       with open(out_file, 'wb') as writer:         # 读取输入的文本文件,使偶数行成为article,奇数行成为abstract(行号从0开始)         lines = read_text_file(input_file)         for i, new_line in enumerate(lines):             if % 2 == 0:                 article = lines[i]             if % 2 != 0:                 abstract = "%s %s %s" % (SENTENCE_START, lines[i], SENTENCE_END)                   # 写入tf.Example                 tf_example = example_pb2.Example()                 tf_example.features.feature['article'].bytes_list.value.extend([bytes(article, encoding='utf-8')])                 tf_example.features.feature['abstract'].bytes_list.value.extend([bytes(abstract, encoding='utf-8')])                 tf_example_str = tf_example.SerializeToString()                 str_len = len(tf_example_str)                 writer.write(struct.pack('q', str_len))                 writer.write(struct.pack('%ds' % str_len, tf_example_str))                   # 如果可以,将词典写入文件                 if makevocab:                     art_tokens = article.split(' ')                     abs_tokens = abstract.split(' ')                     abs_tokens = [t for in abs_tokens if                                   not in [SENTENCE_START, SENTENCE_END]]  # 从词典中删除这些符号                     tokens = art_tokens + abs_tokens                     tokens = [t.strip() for in tokens]     # 去掉句子开头结尾的空字符                     tokens = [t for in tokens if t != ""]  # 删除空行                     vocab_counter.update(tokens)       print("Finished writing file %s " % out_file)       # 将词典写入文件     if makevocab:         print("Writing vocab file...")         with open(os.path.join(FINISHED_FILE_DIR, "vocab"), 'w', encoding='utf-8') as writer:             for word, count in vocab_counter.most_common(VOCAB_SIZE):                 writer.write(word + ' ' + str(count) + ' ')         print("Finished writing vocab file")   if __name__ == '__main__':     if not os.path.exists(FINISHED_FILE_DIR):     os.makedirs(FINISHED_FILE_DIR)     write_to_bin(VAL_FILE, os.path.join(FINISHED_FILE_DIR, "val.bin"))     write_to_bin(TRAIN_FILE, os.path.join(FINISHED_FILE_DIR, "train.bin"), makevocab=True)     chunk_all() 

  在训练中文数据集的时候,设置的hidden_dim为 256 ,词向量维度emb_dim为126,词汇表数目vocab_size为50K,batch_size设为16。这里由于我们的模型有处理OOV能力,因此词汇表不用设置过大;在batch_size的选择上,显存小的同学建议设为8,否则会出现内存不够,难以训练。

  在batch_size=16时,训练了27k step, 出现loss震荡很难收敛的情况,train阶段loss如下:

 

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" width="1050" height="333" /> val阶段loss如下:

  可以看到当step在10k之后,loss在3.0-5.0之间来回剧烈震荡,并没有下降趋势。前面我们为了省显存,将batch_size设置成16,可能有点小了,梯度下降方向不太明确,显得有点盲目,因此将batch_size设成了32后重新开始训练。注意:在一定范围内,batchsize越大,计算得到的梯度下降方向就越准,引起训练震荡越小。增大batch_size后训练的loss曲线如下:

val loss曲线如下:

  看起来loss还是比较震荡的,但是相比bathc_size=16时有所改善。一开始的前10K steps里loss下降还是很明显的基本上能从6降到4左右的区间,10k steps之后开始震荡,但还是能看到在缓慢下降:从4左右,开始在2-4之间震荡下降。这可能是目前的steps还比较少,只要val loss没有一直升高,可以继续观擦,如果500K steps都还是如此,可以考虑在一个合适的实机early stop。

回到顶部

4.3 Evaluation

  摘要质量评价需要考虑一下三点:

    (1) 决定原始文本最重要的、需要保留的部分;

    (2) 在自动文本摘要中识别出1中的部分;

    (3) 基于语法和连贯性(coherence)评价摘要的可读性(readability)。

  从这三点出发有人工评价和自动评价,本文只讨论一下更值得关注的自动评价。自动文档摘要评价方法分为两类:

    内部评价方法(Intrinsic Methods):提供参考摘要,以参考摘要为基准评价系统摘要的质量。系统摘要与参考摘要越吻合, 质量越高。

    外部评价方法(Extrinsic Methods):不提供参考摘要,利用文档摘要代替原文档执行某个文档相关的应用。

  内部评价方法是最常使用的文摘评价方法,将系统生成的自动摘要与参考摘要采用一定的方法进行比较是目前最为常见的文摘评价模式。下面介绍内部评价方法是ROUGE(Recall-Oriented Understudy for Gisting Evaluation)。

  ROUGE是2004年由ISI的Chin-Yew Lin提出的一种自动摘要评价方法,现被广泛应用于DUC(Document Understanding Conference)的摘要评测任务中。ROUGE基于摘要中n元词(n-gram)的共现信息来评价摘要,是一种面向n元词召回率的评价方法。基本思想为由多个专家分别生成人工摘要,构成标准摘要集,将系统生成的自动摘要与人工生成的标准摘要相对比,通过统计二者之间重叠的基本单元(n元语法、词序列和词对)的数目,来评价摘要的质量。通过与专家人工摘要的对比,提高评价系统的稳定性和健壮性。该方法现已成为摘要评价技术的通用标注之一。 ROUGE准则由一系列的评价方法组成,包括ROUGE-N(N=1、2、3、4,分别代表基于1元词到4元词的模型),ROUGE-L,ROUGE-S, ROUGE-W,ROUGE-SU等。在自动文摘相关研究中,一般根据自己的具体研究内容选择合适的ROUGE方法。公式如下: 

 

ROUGEN=S{ReferenceSummaries}gramnSCountmatch(gramn)S{ReferenceSummaries}gramnSCount(gramn)ROUGE−N=∑S∈{ReferenceSummaries}∑gramn∈SCountmatch(gramn)∑S∈{ReferenceSummaries}∑gramn∈SCount(gramn)

 

  其中,ngramn−gram表示n元词RefSummariesRefSummaries表示参考摘要(标准摘要)Countmatch(ngram)Countmatch(n−gram)表示生成摘要和参考摘要中同时出现ngramn−gram的个数Count(ngram)Count(n−gram)则表示参考摘要中出现的ngramn−gram个数ROUGE公式是由召回率的计算公式演变而来的,分子可以看作“检出的相关文档数目”,即系统生成摘要与标准摘要相匹配的ngramn−gram个数,分母可以看作“相关文档数目”,即参考摘要中所有的ngramn−gram个数。

来看原文试验结果:

   在上表中,上半部分是模型生成的的摘要评估,而下半部分的是提取摘要评估。可以看出抽象生成的效果接近了抽取效果。再来看重复情况:

  可以看出我们的no coverage的模型生成的摘要在n-gram上是要比reference摘要要多的,而使用了coverage之后,重复数目和reference相当。 看一下我们的中文结果: 例子一:

例子二:

 

  直观上效果还是不错的。可以看出,预测的摘要中已经基本没有不断重复自身的现象;像“[话筒] [思考] [吃惊] ”这种文本,应该是原文本中的表情,在对文本的处理中我们并没有将这些清洗掉,因此依然出现在预测摘要中。不过例子二还是出现了句子不是很通顺的情况,在输出句子的语序连贯上还有待改进。 回到顶部

4.4 Results

  1. 在复现原论文的基础上,将模型方法应用在中文数据集上,取得了一定效果。

  2. 可以看出指针生成网络通过指针复制原文中的单词,可以生成新的单词,解决oov问题;其次使用了coverage机制,能够避免生成的词语不断重复。

  3. 在语句的通顺和连贯上还有待加强。

回到顶部

5 References

  1. https://arxiv.org/pdf/1704.04368.pdf   2. https://www.jiqizhixin.com/articles/2019-03-25-7   3. https://zhuanlan.zhihu.com/p/53821581   4. https://www.aclweb.org/anthology/W04-1013   5. https://blog.csdn.net/mr2zhang/article/details/90754134   6. https://zhuanlan.zhihu.com/p/68253473

以上就是指针生成网络(Pointer-Generator-Network)原理与实战的全部内容。