代码:https://github.com/asahi417/DocumentClassification

https://github.com/asahi417/DocumentClassification/blob/master/sequence_modeling/model/cnn_char.py

abstract

短文本的情感分析

挑战:有限的文本信息

本文提出DCNN(deep convolutional neural network),从字符级到句子级来进行分析。

语料采用两个不同领域的,Stanford Sentiment Treebank (SSTb),电影评论,Stanford Twitter Sentiment corpus (STS),Twitter信息

1 Introduction

advent: [ˈædˌvent]  n 到来

crescent: [ˈkrez(ə)nt] n 新月形,新月的

 

2 Neural Network Architecture

网络的输入:一句话的词序列,经过一系列层,抽取信息,网络从字符级到句子级抽取信息,

 

2.1 Initial Representation Levels

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

2.1.1 Word-Level Embeddings

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

2.1.2 Character-Level Embeddings

捕获形态和形状的信息,必须考虑词的所有字母和选择哪些信息是重要的。

例如,对Twitter数据的情感分析,重要信息可以出现在#号的部分,“#SoSad”, “#ILikeIt”,或是副词,以“ly”结束,例如“beautifully”, “perfectly” and “badly”,我们使用的方法和dos Santos and Zadrozny, 2014相同,这是基于卷积的方法,具体是:卷积产生词的character-level特征,然后用max-pooling的办法把他们结合在一起,获得这个词的基于character-level的词嵌入。

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

2.2 Sentence-Level Representation and Scoring

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

2.3 Network Training

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

3 Related Work

关于情感分析的网络结构

循环网络

Socher et al., 2011:a semi-supervised approach based on recursive autoencoders for predicting sentiment distributionsè这个方法给不同的短语学习向量空间表示,并探索句子的递归特征。

Socher et al., 2012:a matrix-vector recursive neural network model for semantic compositionalityè有能力学习短语和句子的compositional 向量表示。这些向量捕获了潜在句子结构,矩阵捕获邻近的词和短语的意思。

Socher et al., 2013b:Recursive Neural Tensor Network (RNTN) architectureè用词向量表示短语和parse tree,然后计算高节点的向量。

我们的工作采用前向传播网络,而不是递归的。

NLP’任务中的卷积网络

Collobert et al., 2011:use a convolutional network for the semantic role labeling task with the goal avoiding excessive task-specific feature engineering

Collobert, 2011:use a similar network architecture for syntactic parsing

CharSCNN与这些网络有关,使用一层卷积层提取句子特征,这些网络的不同在于使用多一层卷积层进行字符级的特征提取。

在神经网络结构使用intra-word,

Luong et al., 2013:use a recursive neural network (RNN) to explicitly model the morphological structures of words and learn morphologically-aware embeddings

Lazaridou et al., 2013:use compositional distributional semantic models, originally designed to learn meanings of phrases, to derive representations for complex words, in which the base unit is the morpheme

Chrupala, 2013:proposes a simple recurrent network (SRN) to learn continuous vector representations for sequences of characters, and use them as features in a conditional random field classifier to solve a character level text segmentation and labeling task

4 Experimental Setup and Results

4.1 Sentiment Analysis Datasets

(Stanford Sentiment Treebank (SSTb)

Stanford Twitter Sentiment corpus (STS):test set was manually annotated by Go et al

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

 

4.2 Unsupervised Learning of Word-Level Embeddings

word-level的embedding对于CharSCNN结构很重要,他们是要捕获syntactic和semantic信息,这对于情感分析特别重要,这些工作可以有下成果获得:

Collobert et al., 2011; Luong et al., 2013; Zheng et al., 2013; Socher et al., 2013a

Mikolov et al., 2013:implements the continuous bag-of-words and skip-gram architectures for computing vector representations of words

 

 

4.3 Model Setup

 

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

 

 

4.4 Results for SSTb Corpus

 

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

 

4.5 Results for STS Corpus

 

Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

 

4.6 Sentence-level features

 

5 Conclusions

 

 

 

相关文章: