文本情感分析综述

以文本颗粒度为视角,从情感词抽取、语料库和情感词典构建、评价对象与意见持有者分析、篇章级情感分析、实际应用五个方面对文本情感分析文献进行了梳理,并做出必要评述。指出当前情感分析系统的准确率普遍不高,进一步研究的重点在于：自然语言处理的研究成果在文本情感倾向分析中更广泛和贴切的应用;选取文本情感倾向分类的特征和方法;利用现有语言工具和相关资源,规范、快速地构造语言工具和相关资源并应用。     

Building and evaluating resources for sentiment analysis in the Greek language

Sentiment lexicons and word embeddings constitute well-established sources of information for sentiment analysis in online social media. Although their effectiveness has been demonstrated in state-of-the-art sentiment analysis and related tasks in the English language, such publicly available resources are much less developed and evaluated for the Greek language. In this paper, we tackle the problems arising when analyzing text in such an under-resourced language. We present and make publicly available a rich set of such resources, ranging from a manually annotated lexicon, to semi-supervised word embedding vectors and annotated datasets for different tasks. Our experiments using different algorithms and parameters on our resources show promising results over standard baselines; on average, we achieve a 24.9% relative improvement in F-score on the cross-domain sentiment analysis task when training the same algorithms with our resources, compared to training them on more traditional feature sources, such as n-grams. Importantly, while our resources were built with the primary focus on the cross-domain sentiment analysis task, they also show promising results in related tasks, such as emotion analysis and sarcasm detection.     

情感倾向性分析及应用研究综述

在情感倾向性分析领域,关于情感的收集、分析和聚合等技术一直是近年来的关注热点。该领域的相关发展带动了各个子任务及其相关研究的大力发展。本文主要综述了面向情感的信息系统中使用的情感分析相关的需求,技术,应用以及评测方法等。在情报分析方面,存在许多不同于传统的主题分析的新需求,这就是对情感分析技术的强烈需求。接着,介绍了词级、句子级、段落篇章级等不同层次的情感分析技术。然后,还综述了采用情感分析技术的各种典型应用。最后,为了工作开展的便利,讨论了情感分析领域的词库资源、样本集资源、评测方法及重要会议等。     

基于多通道特征和自注意力的情感分类方法

针对情感分析任务中没有充分利用现有的语言知识和情感资源,以及在序列模型中存在的问题：模型会将输入文本序列解码为某一个特定的长度向量,如果向量的长度设定过短,会造成输入文本信息丢失.提出了一种基于多通道特征和自注意力的双向LSTM情感分类方法（MFSA-BiLSTM）,该模型对情感分析任务中现有的语言知识和情感资源进行建模,形成不同的特征通道,并使用自注意力重点关注加强这些情感信息.MFSA-BiLSTM可以充分挖掘句子中的情感目标词和情感极性词之间的关系,且不依赖人工整理的情感词典.另外,在MFSA-BiLSTM模型的基础上,针对文档级文本分类任务提出了MFSA-BiLSTM-D模型.该模型先训练得到文档的所有的句子表达,再得到整个文档表示.最后,对5个基线数据集进行了实验验证.结果表明：在大多数情况下,MFSA-BiLSTM和MFSA-BiLSTM-D这两个模型在分类精度上优于其他先进的文本分类方法.     

基于双语信息和标签传播算法的中文情感词典构建方法

文本情感分析是目前自然语言处理领域的一个热点研究问题,具有广泛的实用价值和理论研究意义。情感词典构建则是文本情感分析的一项基础任务,即将词语按照情感倾向分为褒义、中性或者贬义。然而,中文情感词典构建存在两个主要问题 1)许多情感词存在多义、歧义的现象,即一个词语在不同语境中它的语义倾向也不尽相同,这给词语的情感计算带来困难;2)由国内外相关研究现状可知,中文情感字典建设的可用资源相对较少。考虑到英文情感分析研究中存在大量语料和词典,该文借助机器翻译系统,结合双语言资源的约束信息,利用标签传播算法(LP)计算词语的情感信息。在四个领域的实验结果显示我们的方法能获得一个分类精度高、覆盖领域语境的中文情感词典。     

SlangSD: building,expanding and using a sentiment dictionary of slang words for short-text sentiment classification

Sentiment information about social media posts is increasingly considered an important resource for customer segmentation, market understanding, and tackling other socio-economic issues. However, sentiment in social media is difficult to measure since user-generated content is usually short and informal. Although many traditional sentiment analysis methods have been proposed, identifying slang sentiment words remains a challenging task for practitioners. Though some slang words are available in existing sentiment lexicons, with new slang being generated with emerging memes, a dedicated lexicon will be useful for researchers and practitioners. To this end, we propose to build a slang sentiment dictionary to aid sentiment analysis. It is laborious and time-consuming to collect a comprehensive list of slang words and label the sentiment polarity. We present an approach to leverage web resources to construct a Slang Sentiment Dictionary (SlangSD) that is easy to expand. SlangSD is publicly available for research purposes. We empirically show the advantages of using SlangSD, the newly-built slang sentiment word dictionary for sentiment classification, and provide examples demonstrating its ease of use with a sentiment analysis system.     

基于多通道双向长短期记忆网络的情感分析

当前存在着大量的语言知识和情感资源,但在基于深度学习的情感分析研究中,这些特有的情感信息,没有在情感分析任务中得到充分利用。针对以上问题,该文提出了一种基于多通道双向长短期记忆网络的情感分析模型(multi-channels bidirectional long short term memory network,Multi-Bi-LSTM),该模型对情感分析任务中现有的语言知识和情感资源进行建模,生成不同的特征通道,让模型充分学习句子中的情感信息。与CNN相比,该模型使用的Bi-LSTM考虑了词序列之间依赖关系,能够捕捉句子的上下文语义信息,使模型获得更多的情感信息。最后在中文COAE2014数据集、英文MR数据集和SST数据集进行实验,取得了比普通Bi-LSTM、结合情感序列特征的卷积神经网络以及传统分类器更好的性能。     

Is there a language of sentiment? An analysis of lexical resources for sentiment analysis

In recent years, sentiment analysis (SA) has emerged as a rapidly expanding field of application and research in the area of information retrieval. In order to facilitate the task of selecting lexical resources for automated SA systems, this paper sets out a detailed analysis of four widely used sentiment lexica. The analysis provides an overview of the coverage of each lexicon individually, the overlap and consistency of the four resources and a corpus analysis of the distribution of the resources’ lexical contents in general and specialised language. This work aims to explore the characteristics of affective language as represented by these lexica and the implications of the findings for developers of SA systems.     

结合卷积神经网络和词语情感序列特征的中文情感分析

目前基于词嵌入的卷积神经网络文本分类方法已经在情感分析研究中取得了很好的效果。此类方法主要使用基于上下文的词嵌入特征,但在词嵌入过程中通常并未考虑词语本身的情感极性,同时此类方法往往缺乏对大量人工构建情感词典等资源的有效利用。针对这些问题,该文提出了一种结合情感词典和卷积神经网络的情感分类方法,利用情感词典中的词条对文本中的词语进行抽象表示,在此基础上利用卷积神经网络提取抽象词语的序列特征,并用于情感极性分类。该文提出的相关方法在中文倾向性分析评测COAE2014数据集上取得了比目前主流的卷积神经网络以及朴素贝叶斯支持向量机更好的性能。     

跨领域文本情感分类研究进展

作为社会媒体文本情感分析的重要研究课题之一，跨领域文本情感分类旨在利用源领域资源或模型迁移地服务于目标领域的文本情感分类任务，其可以有效缓解目标领域中带标签数据不足问题.本文从三个角度对跨领域文本情感分类方法行了归纳总结：（1）按照目标领域中是否有带标签数据，可分为直推式和归纳式情感迁移方法；（2）按照不同情感适应性策略，可分为实例迁移方法、特征迁移方法、模型迁移方法、基于词典的方法、联合情感主题方法以及图模型方法等；（3）按照可用源领域个数，可分为单源和多源跨领域文本情感分类方法.此外，论文还介绍了深度迁移学习方法及其在跨领域文本情感分类的最新应用成果.最后，论文围绕跨领域文本情感分类面临的关键技术问题，对可能的突破方向进行了展望.     

基于LSTM的商品评论情感分析

随着电子商务的发展,产生了大量的商品评论文本.针对商品评论的短文本特征,基于情感词典的情感分类方法需要大量依赖于情感数据库资源,而机器学习的方法又需要进行复杂的人工设计特征和提取特征过程.本文提出采用长短期记忆网络（Long Short-Term Memory）文本分类算法进行情感倾向分析,首先利用Word2vec和分词技术将评论短文本文本处理为计算机可理解的词向量传入LSTM网络并加入Dropout算法以防止过拟合得出最终的分类模型.实验表明：在基于深度学习的商品评论情感倾向分析中,利用LSTM网络的短时记忆独特特征对商品评论的情感分类取得了很好的效果,准确率达到99%以上.     

基于词典和规则集的中文微博情感分析

通过对微博文本的特性分析,提取了中文微博情感分析的关键问题：如何识别微博新词并理解其情感含义?如何利用附加信息辅助文本情感分析?如何结合语言特性构造情感计算方法?针对第一个问题,利用统计信息和点间互信息对新词进行挖掘和情感识别,在40万条新浪微博数据中构建了新情感词词典,用于对已有情感词资源的扩充。对于后两个问题,提出了基于词典和规则集的中文微博情感分析方法。根据微博特性,在不同的语言层次上定义了规则,结合情感词典对微博文本进行了从词语到句子的多粒度情感计算,并以表情符号作为情感计算的辅助元素。通过对采集到的原创微博数据集进行实验,验证了该方法的有效性。     

Building layered,multilingual sentiment lexicons at synset and lemma levels

Many tasks related to sentiment analysis rely on sentiment lexicons, lexical resources containing information about the emotional implications of words (e.g., sentiment orientation of words, positive or negative). In this work, we present an automatic method for building lemma-level sentiment lexicons, which has been applied to obtain lexicons for English, Spanish and other three official languages in Spain. Our lexicons are multi-layered, allowing applications to trade off between the amount of available words and the accuracy of the estimations. Our evaluations show high accuracy values in all cases. As a previous step to the lemma-level lexicons, we have built a synset-level lexicon for English similar to SentiWordNet 3.0, one of the most used sentiment lexicons nowadays. We have made several improvements in the original SentiWordNet 3.0 building method, reflecting significantly better estimations of positivity and negativity, according to our evaluations. The resource containing all the lexicons, ML-SentiCon, is publicly available.     

SentiTurkNet: a Turkish polarity lexicon for sentiment analysis

Sentiment analysis aims to extract the sentiment polarity of given segment of text. Polarity resources that indicate the sentiment polarity of words are commonly used in different approaches. While English is the richest language in regard to having such resources, the majority of other languages, including Turkish, lack polarity resources. In this work we present the first comprehensive Turkish polarity resource, SentiTurkNet, where three polarity scores are assigned to each synset in the Turkish WordNet, indicating its positivity, negativity, and objectivity (neutrality) levels. Our method is general and applicable to other languages. Evaluation results for Turkish show that the polarity scores obtained through this method are more accurate compared to those obtained through direct translation (mapping) from SentiWordNet.     

评价对象抽取研究综述

近年来,细粒度情感分析因其在商业决策、舆情分析等领域的重要作用而受到学术界和工业界的广泛关注.评价对象抽取作为情感分析的基本任务之一,是进行细粒度情感分析的关键问题.本文针对评价对象抽取问题的起源、当前主流研究方法和趋势进行了梳理,首先详细阐述评价对象抽取问题的基本概念并对其进行形式化表示,然后结合近年来的研究对评价对象抽取方法进行归纳和总结,并重点分析基于频率、基于模板规则、基于图论、基于条件随机场和基于深度学习的评价对象抽取方法,随后回顾评价对象抽取的评测情况和可用的语料资源,最后分析评价对象抽取的若干难点问题,同时对评价对象抽取研究进展和发展趋势进行总结和展望.     

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％.     

基于领域情感词典特征表示的细粒度意见挖掘

细粒度意见挖掘的主要目标是从观点文本中获取情感要素并判断情感倾向。现有方法大多基于序列标注模型,但很少利用情感词典资源。该文提出一种基于领域情感词典特征表示的细粒度意见挖掘方法,使用领域情感词典在观点文本上构建特征表示并将其加入序列标注模型的输入部分。首先构建一份新的电商领域情感词典,然后在电商评论文本真实数据上,分别为条件随机场(CRF)和双向长短期记忆-条件随机场(BiLSTM-CRF)这两种常用序列标注模型设计基于领域情感词典的特征表示。实验结果表明,基于电商领域情感词典的特征表示方法在两种模型上都取得了良好的效果,并且超过其他情感词典。     

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.     

Comparative experiments using supervised learning and machine translation for multilingual sentiment analysis

Sentiment analysis is the natural language processing task dealing with sentiment detection and classification from texts. In recent years, due to the growth in the quantity and fast spreading of user-generated contents online and the impact such information has on events, people and companies worldwide, this task has been approached in an important body of research in the field. Despite different methods having been proposed for distinct types of text, the research community has concentrated less on developing methods for languages other than English. In the above-mentioned context, the present work studies the possibility to employ machine translation systems and supervised methods to build models able to detect and classify sentiment in languages for which less/no resources are available for this task when compared to English, stressing upon the impact of translation quality on the sentiment classification performance. Our extensive evaluation scenarios show that machine translation systems are approaching a good level of maturity and that they can, in combination to appropriate machine learning algorithms and carefully chosen features, be used to build sentiment analysis systems that can obtain comparable performances to the one obtained for English.     

A cross-media public sentiment analysis system for microblog

Since classical public sentiment analysis systems for microblog are based on the text sentiment analysis, it is difficult to determine the sentiment of short text without clear sentiment words in microblog posts. Fortunately, a lot of microblog posts contain images which also represent users’ sentiment. To fully understand users’ sentiment, we propose a cross-media public sentiment analysis system for microblog. The best advantage of this novel system is the unified cross-media public sentiment analysis framework which fuses the text sentiment and image sentiment not only from sentiment results, but also from sentiment ontology. To enhance presentation effects, this system presents sentiment results from macroscopic view and microscopic view which details the sentiment results in region, topic, microblog content and user diffusion. In our knowledge, this is the first unified cross-media public sentiment analysis system.