面向中日关系论坛的情感分类问题研究

针对中日论坛的情感分类问题,研究了特定领域（中日关系论坛）语料的特点,考察了不同特征维数、不同特征权重计算、不同特征选取方法以及限定词类词语对情感分类结果的影响。最后通过对2006年1月份到5月份的中日论坛语料的自动情感分类,推断出该阶段中日关系走势。     

结合学科情感分析与依存关系的相似度评分

通过对语文古诗文阅读类主观题的分析,提出了结合学科情感分析与依存关系的相似度评分算法,并将其应用于高中语文古诗文阅读类主观题的评分中.首先,以中文维基百科语料为基础,增加了与评分相关的古诗文语料81927条,通过文本向量化算法Word2vec进行词向量训练,完成了对古诗文语料库的构建;基于学科评分特性建立了对应的古诗文...     

Visual Analysis of Sentiment and Information Spread on Micro-Blog

In this paper, we combined temporal analysis and spatial analysis together, and proposed the Electron Cloud Model (ECM) which is based on the Schrodinger equation and Niels Bohr atomic theory. The ECM is used to conduct temporal visual analysis of micro-blog sentiments. In the ECM, we made an attempt to mapping a score of sentiment to the electron stability and took neutral sentiments into consideration. We applied kernel density estimation and edge bundling to conduct space-varying visual analysis of sentiment. Kernel density estimation visualized sentiment changes in different levels of detail naturally while edge bundling was used to reduce visual clutter of edge crossing and reveal high-level edge patterns. Finally, we implemented an analysis system, conducted three case studies and made simple comparisons with other visualize methods.     

基于CRT机制混合神经网络的特定目标情感分析

特定目标情感分析的目的是从不同目标词语的角度来预测文本的情感,关键是为给定的目标分配适当的情感词。当句子中出现多个情感词描述多个目标情感的情况时,可能会导致情感词和目标之间的不匹配。由此提出了一个CRT机制混合神经网络用于特定目标情感分析,模型使用CNN层从经过BiLSTM变换后的单词表示中提取特征,通过CRT组件生成单词的特定目标表示并保存来自BiLSTM层的原始上下文信息。在三种公开数据集上进行了实验,结果表明,该模型在特定目标情感分析任务中较之前的情感分析模型在准确率和稳定性上有着明显的提升,证明CRT机制能很好地整合CNN和LSTM的优势,这对于特定目标情感分析任务具有重要的意义。     

Domain-specific sentiment analysis for tweets during hurricanes (DSSA-H): A domain-adversarial neural-network-based approach

Hurricanes are one of the most frequent and destructive disasters in the United States. The events are large scale and have relatively long-term impacts. Social networking platforms such as Twitter can provide real-time information for disaster managers and affected populations during large-scale disasters (e.g., hurricanes), but extracting useful information and interpreting data accurately for disaster management is still challenging. Sentiment analysis of social media data helps detect the concerns of affected people and understand individuals' responses on the ground at unprecedented scales, but the method is known to be domain-dependent. The same words or expressions can indicate opposite sentiments in different domains. This paper proposes a domain-specific sentiment analysis approach specifically for tweets posted during hurricanes (DSSA-H). DSSA-H can retrieve hurricane-relevant tweets with a trained supervised-learning classifier, Random Forest (RF), and classify the sentiment of hurricane-relevant tweets based on a domain-adversarial neural network (DANN). We built a dataset of tweets posted during six recent hurricanes and applied the DSSA-H approach for sentiment analysis. After evaluation, we found that each classifier (i.e., RF and DANN) outperforms baseline classifiers and that DSSA-H outperforms two high-performing general sentiment classification approaches when classifying sentiments of tweets posted during hurricanes. We also applied DSSA-H in examining sentiment patterns across six recent hurricanes in the U.S. This domain-specific sentiment analysis approach can be used by the first responders and affected communities to more accurately and rapidly detect crises and emergent events, allocate resources, and assess disaster's impact during hurricanes. DSSA-H contributes to an intelligent and adaptive disaster information system for the data-rich human and the built environment system.     

基于依存句法的博文情感倾向分析研究

博客作为一种用户发表其观点和看法的载体已成为Web上一个重要的情感抒发与交流平台,博文搜索为这种交流提供了方便快捷的途径.很多时候,用户进行博文搜索时更关注作者对事件所持的观点或情感,但目前的博文搜索返回结果大多基于主题而非情感倾向.基于此提出一种基于句法依存分析技术的算法SOAD(sentiment orientation analysis based on syntactic dependency)对博文搜索结果进行情感倾向性分析.基于SOAD算法,构建了一个中文博文搜索原型系统,对博文搜索结果进行再处理.实验证明,一方面,SOAD算法在分析博文情感上具有更大的优势;另一方面,建立的原型系统实现了依据情感倾向返回搜索结果的目标.     

基于神经网络融合标签相关性的多标签情感预测研究

近年来,多标签分类任务(MLC)受到了广泛关注。传统的情感预测被视为一种单标签的监督学习,而忽视了多种情感可能在同一实例中共存的问题。以往的多标签情感预测方法没有同时提取文本的局部特征和全局语义信息,或未考虑标签之间的相关性。基于此,该文提出了一种基于神经网络融合标签相关性的多标签情感预测模型(Label-CNNLSTMAttention,L-CLA),利用Word2Vec方法训练词向量,将CNN和LSTM相结合,通过CNN层挖掘文本更深层次的词语特征,通过LSTM层学习词语之间的长期依赖关系,利用Attention机制为情意词特征分配更高的权重。同时,用标签相关矩阵将标签特征向量补全后与文本特征共同作为分类器的输入,考察了标签之间的相关性。实验结果表明,L-CLA模型在重新标注后的NLP&CC2013数据集上拥有较好的分类效果。     

ABAFN: 面向多模态的方面级情感分类模型

通过方面术语提取和方面级情感分类任务提取句子中的方面-情感对,有助于Twitter,Facebook等社交媒体平台挖掘用户对不同方面的情感,对个性化推荐有重要的意义. 在多模态领域,现有方法使用2个独立的模型分别完成2个子任务,方面术语提取提取句子中包含的商品、重要人物等实体或实体的方面,方面级情感分类根据给定的方面术语预测用户的情感倾向. 上述方法存在2个问题：1）使用2个独立的模型丢失了2个任务之间在底层特征的延续性,无法建模句子潜在的语义关联;2）方面级情感分类1次预测1个方面的情感,与方面术语提取同时提取多个方面的吞吐量不匹配,且2个模型串行执行使得提取方面-情感对的效率低. 为解决这2个问题,提出基于多模态方面术语提取和方面级情感分类的统一框架UMAS. 首先,建立共享特征模块,实现任务间潜在语义关联建模,并且共享表示层使得2个子任务只需关心各自上层的网络,降低了模型的复杂性;其次,模型利用序列标注同时输出句子中包含的多个方面及其对应的情感类别,提高了方面-情感对的提取效率. 此外,在这2个子任务中同时引入词性：利用其中蕴含的语法信息提升方面术语提取的性能;通过词性获取观点词信息,提升方面级情感分类的性能. 实验结果表明,该统一框架在Twitter2015,Restaurant2014这2个基准数据集上相比于多个基线模型具有优越的性能.

     

维吾尔文情感语料库的构建与分析

针对维吾尔文情感语料库标注体系不规范、语料库规模小、没有合适的标注平台等问题,分析英文和中文比较著名情感语料库的优点,结合维吾尔语文本的特点,建立维吾尔文情感语料标注规范,利用Python语言构建集数据采集与标注为一体的情感标注平台,最后构建在舆情分析和舆情监控中可以应用的维吾尔文情感语料库。实验结果表明,该标注规范具有可扩展性和实用性,标注平台可以有效地减轻标注人员的工作量,提高情感语料库的质量,情感语料库可以用于舆情分析任务。      

Sentiment community detection: exploring sentiments and relationships in social networks

Social networking sites (SNS), which allow users to express opinions on products/services, have become an important channel and platform for enterprises to acquire and trace users’ sentiments in order to design appropriate business strategies and online marketing campaigns. However, with the large number of users and complex user relationships on SNS, effectively capturing these sentiments for business decision support is still a big challenge. In this study we introduce the concept of “Sentiment Community,” a group of users who are closely connected and highly consistent in their sentiments about one product/service. Discovering such sentiment communities would be very valuable to enterprises for customer segmentation and target marketing. Taking into account both connections and sentiments, we propose two methods to discover sentiment communities by adopting the optimization models of semi-definite programming (SDP). Our experimental evaluations demonstrated great performances for the proposed methods. This study opens the doors to effectively explore users’ sentiments on SNS for business decision making.     

Sentiment analysis leveraging emotions and word embeddings

Sentiment analysis and opinion mining are valuable for extraction of useful subjective information out of text documents. These tasks have become of great importance, especially for business and marketing professionals, since online posted products and services reviews impact markets and consumers shifts. This work is motivated by the fact that automating retrieval and detection of sentiments expressed for certain products and services embeds complex processes and pose research challenges, due to the textual phenomena and the language specific expression variations. This paper proposes a fast, flexible, generic methodology for sentiment detection out of textual snippets which express people’s opinions in different languages. The proposed methodology adopts a machine learning approach with which textual documents are represented by vectors and are used for training a polarity classification model. Several documents’ vector representation approaches have been studied, including lexicon-based, word embedding-based and hybrid vectorizations. The competence of these feature representations for the sentiment classification task is assessed through experiments on four datasets containing online user reviews in both Greek and English languages, in order to represent high and weak inflection language groups. The proposed methodology requires minimal computational resources, thus, it might have impact in real world scenarios where limited resources is the case.     

A novel approach for clustering sentiments in Chinese blogs based on graph similarity

Blog clustering is an important approach for online public opinion analysis. The traditional clustering methods, usually group blogs by keywords, stories and timeline, which usually ignore opinions and emotions expressed in the blog articles. In this paper, an integrated graph-based model for clustering Chinese blogs by embedded sentiments is proposed. A novel graph-based representation and the corresponding clustering algorithm are applied on the Chinese blog search results. The proposed model SoB-graph considers not only sentiment words but also structural information in blogs. Experimental results show that comparing with the traditional graph-based document representation model and vector space document representation model, the proposed SoB-graph model has achieved better performance in clustering sentiments in Chinese blog documents.     

基于深度半监督的柬语句子级情感分类

针对柬语标注数据较少、语料稀缺,柬语句子级情感分析任务进步缓慢的问题,提出了一种基于深度半监督CNN(convolutional neural networks)的柬语句子级情感极性分类方法.该方法通过融合词典嵌入的分开卷积CNN模型,利用少量已有的柬语情感词典资源提升句子级情感分类任务性能.首先构建柬语句子词嵌入和词典嵌入,通过使用不同的卷积核对两部分嵌入分别进行卷积,将已有情感词典信息融入到CNN模型中去,经过最大延时池化得到最大输出特征,把两部分最大输出特征拼接后作为全连接层输入;然后通过结合半监督学习方法——时序组合模型,训练提出的深度神经网络模型,利用标注与未标注语料训练,降低对标注语料的需求,进一步提升模型情感分类的准确性.结果 证明,通过半监督方法时序组合模型训练,在人工标记数据相同的情况下,该方法相较于监督方法在柬语句子级情感分类任务上准确率提升了3.89％.     

面向商品评论的二元情感认知模型

该文提出了一种面向商品评论的二元情感认知模型。该模型由“二元情感常识库”、“评价体系知识库”和“情感分析引擎”三个主要模块组成。其特点体现为:(1)模型通过大规模评论文本学习领域先验知识,将其存储在知识库中,便于知识的修正和重用,体现了模型的认知能力;(2)模型不仅能够挖掘评论文本中出现的显式评价观点,还能借助领域知识进行情感推断,发现更高层次的用户情感。该文给出了构建“二元情感常识库”和“评价体系知识库”的相关算法,并介绍了“情感分析引擎”在观点挖掘和情感推断中的应用。在商品评论语料集上的实验验证了该模型的有效性。     

A decision-making algorithm for online shopping using deep-learning–based opinion pairs mining and q-rung orthopair fuzzy interaction Heronian mean operators

In the process of online shopping, consumers usually compare the review information of the same product in different e-commerce platforms. The sentiment orientation of online reviews from different platforms interactively influences on consumers’ purchase decision. However, due to the limitation of the ability to process information manually, it is difficult for a consumer to accurately identify the sentiment orientation of all reviews one by one and describe the process of their interactive influence. To this end, we proposed an online shopping support model using deep-learning–based opinion mining and q-rung orthopair fuzzy interaction weighted Heronian mean (q-ROFIWHM) operators. First, in the proposed method, the deep-learning model is used to automatically extract different product attribute words and opinion words from online reviews, and match the corresponding attribute-opinion pairs; meanwhile, the sentiment dictionary is used to calculate sentiment orientation, including positive, negative, and neutral sentiments. Second, the proportions of the three kinds of sentiments about each attribute of the same product are calculated. According to the proportion value of attribute sentiment from different platforms, the sentiment information is converted into multiple cross-decision matrices, which are represented by the q-rung orthopair fuzzy set. Third, considering the interactive characteristics of decision matrix, the q-ROFIWHM operators are proposed to aggregate this cross-decision information, and then the ranking result was determined by score function to support consumers' purchase decisions. Finally, an actual example of mobile phone purchase is given to verify the rationality of the proposed method, and the sensitivity and the comparison analysis are used to show its effectiveness and superiority.     

一种基于情感句模的文本情感分类方法

考虑到同类型的情感句往往具有相同或者相似的句法和语义表达模式,该文提出了一种基于情感句模的文本情感自动分类方法。首先,将情感表达相关句模人工分为3大类105个二级分类;然后,设计了一种利用依存特征、句法特征和同义词特征的句模获取方法,从标注情感句中半自动地获取情感句模。最后,通过对输入句进行情感句模分类实现文本情感分类。在NLP&CC2013中文微博情绪分类评测语料及RenCECps博客语料的实验结果显示,该文提出的分类方法准确率显著高于基于词特征支持向量机分类器。     

基于粗糙集和多通道词向量的中文文本情感特征分析

粗糙集是一种能够有效处理不精确、不完备和不确定信息的数学工具,粗糙集的属性约简可以在保持文本情感分类能力不变的情况下对文本情感词特征进行约简。针对情感词特征空间维数过高、情感词特征表示缺少语义信息的问题,该文提出了RS-WvGv中文文本情感词特征表示方法。利用粗糙集决策表对整个语料库进行情感词特征建模,采用Johnson粗糙集属性约简算法对决策表进行化简,保留最小的文本情感词特征属性集,之后再对该集合中的所有情感特征词进行词嵌入表示,最后用逻辑回归分类器验证RS-WvGv方法的有效性。另外,该文还定义了情感词特征属性集覆盖力,用于表示文本情感词特征属性集合对语料库的覆盖能力。最后,在实验对比的过程中,用统计检验进一步验证了该方法的有效性。     

The hasty wisdom of the mob: How market sentiment predicts stock market behavior

We explore the ability of sentiment metrics, extracted from micro-blogging sites, to predict stock markets. We also address sentiments’ predictive time-horizons. The data concern bloggers’ feelings about five major stocks. Taking independent bullish and bearish sentiment metrics, granular to two minute intervals, we model their ability to forecast stock price direction, volatility, and traded volume. We find evidence of a causal link from sentiments to stock price returns, volatility and volume. The predictive time-horizon is minutes, rather than hours or days. We argue that diverse and high volume sentiment is more predictive of price volatility and traded volume than near-consensus is predictive of price direction. Causality is ephemeral. In this sense, the crowd is more a hasty mob than a source of wisdom.     

Biometric signature verification system based on freeman chain code and k-nearest neighbor

Signature is one of human biometrics that may change due to some factors, for example age, mood and environment, which means two signatures from a person cannot perfectly matching each other. A Signature Verification System (SVS) is a solution for such situation. The system can be decomposed into three stages: data acquisition and preprocessing, feature extraction and verification. This paper presents techniques for SVS that uses Freeman chain code (FCC) as data representation. Before extracting the features, the raw images will undergo preprocessing stage; binarization, noise removal, cropping and thinning. In the first part of feature extraction stage, the FCC was extracted by using boundary-based style on the largest contiguous part of the signature images. The extracted FCC was divided into four, eight or sixteen equal parts. In the second part of feature extraction, six global features were calculated against split image to test the feature efficiency. Finally, verification utilized Euclidean distance to measured and matched in k-Nearest Neighbors. MCYT bimodal database was used in every stage in the system. Based on the experimental results, the lowest error rate for FRR and FAR were 6.67 % and 12.44 % with AER 9.85 % which is better in term of performance compared to other works using that same database.     

多目标情感分类中文数据集构建及分析研究

目标级情感分类任务是为了得到句子中特定评价目标的情感倾向.一个句子中往往存在多个目标,多个目标的情感可能一致,也可能不一致.但在已有针对目标级情感分类的评测数据集中:①大多数是一个句子一个目标;②在少数有多个目标的句子中,多个目标情感倾向分布并不均衡,多个目标情感一致的句子占较大比例.数据集本身的缺陷限制了模型针对多个...