Event extraction using behaviors of sentiment signals and burst structure in social media

Significant world events often cause the behavioral convergence of the expression of shared sentiment. This paper examines the use of the blogosphere as a framework to study user psychological behaviors, using their sentiment responses as a form of ‘sensor’ to infer real-world events of importance automatically. We formulate a novel temporal sentiment index function using quantitative measure of the valence value of bearing words in blog posts in which the set of affective bearing words is inspired from psychological research in emotion structure. The annual local minimum and maximum of the proposed sentiment signal function are utilized to extract significant events of the year and corresponding blog posts are further analyzed using topic modeling tools to understand their content. The paper then examines the correlation of topics discovered in relation to world news events reported by the mainstream news service provider, Cable News Network, and by using the Google search engine. Next, aiming at understanding sentiment at a finer granularity over time, we propose a stochastic burst detection model, extended from the work of Kleinberg, to work incrementally with stream data. The proposed model is then used to extract sentimental bursts occurring within a specific mood label (for example, a burst of observing ‘shocked’). The blog posts at those time indices are analyzed to extract topics, and these are compared to real-world news events. Our comprehensive set of experiments conducted on a large-scale set of 12 million posts from Livejournal shows that the proposed sentiment index function coincides well with significant world events while bursts in sentiment allow us to locate finer-grain external world events.     

基于稀疏组lasso-granger因果关系特征的脑电信号情感识别

针对目前基于单一脑区功能性网络层面的特征提取,文中提出稀疏组lasso-granger因果关系方法.首先从效应性脑网络层面提取不同脑区之间的因果关系作为脑电特征,分别提取受试者α,β,γ脑电波段的granger因果特征值.然后引用稀疏组lasso算法对获取的granger因果特征值进行特征筛选,获得高相关性特征子集作为情感分类特征.最后使用SVM分类器进行情感分类.此外,为了减少计算时间复杂度,使用过滤特征选择(ReliefF)算法,选取有效的脑电信号通道.实验表明,文中方法在Valence-Arousal二维情感模型上获得较高的平均情感分类准确率,分类效果优于对比的脑电特征,提取的情感脑电特征可以有效识别受试者的不同情感状态.     

基于情感向量空间模型的歌曲情感标签预测模型

音乐的情感标签预测对音乐的情感分析有着重要的意义。该文提出了一种基于情感向量空间模型的歌曲情感标签预测算法,首先,提取歌词中的情感特征词构建情感空间向量模型,然后利用SVM分类器对已知情感标签的音乐进行训练,通过分类技术找到与待预测歌曲情感主类一致的歌曲集合,最后,通过歌词的情感相似度计算找到最邻近的k首歌曲,将其标签推荐给待预测歌曲。实验发现本文提出的情感向量空间模型和“情感词—情感标签”共现的特征降维方法比传统的文本特征向量模型能够更好地提高歌曲情感分类准确率。同时,在分类基础上进行的情感标签预测方法可以有效地防止音乐“主类情感漂移”,比最近邻居方法达到更好的标签预测准确率。     

基于脑功能网络和共空间模式分析脑电情绪识别

传统脑网络的情绪分类将聚类系数、平均最短路径等拓扑属性作为分类特征。针对这些属性易受网络连接阈值和特征选择的影响,难以完全表征不同情绪状态下的网络空间拓扑结构差异的问题,提出了一种基于脑网络和共空间模式的脑电情绪识别方法(EEG emotion classification based on common spatial patterns of brain networks topology,EEC-CSP-BNT)。该算法基于互信息在各个子频段内计算电极间的功能连接矩阵,同时利用共空间模式(common spatial pattern,CSP)分析学习空间滤波器,构建分类特征,最后通过分类器(如Fisher线性判别、支持向量机、K最近邻)实现基于脑电的情绪分类。基于DEAP和SEED数据集的实验结果表明,相比于脑网络拓扑属性,EEC-CSP-BNT能有效提取脑网络拓扑结构的分类信息,提高脑电情绪识别性能。     

基于邻域粗糙集的大规模层次分类在线流特征选择

在分类学习任务中,数据的类标记空间存在层次化结构,特征空间伴随着未知性和演化性.因此,文中提出面向大规模层次分类学习的在线流特征选择框架.定义面向层次化结构数据的邻域粗糙模型,基于特征相关性进行重要特征动态选择.最后,基于特征冗余性进行鉴别冗余动态特征.实验验证文中算法的有效性.     

An approach of genetic programming for music emotion classification

In this paper, we suggest a new approach of genetic programming for music emotion classification. Our approach is based on Thayer’s arousal-valence plane which is one of representative human emotion models. Thayer’s plane which says human emotions is determined by the psychological arousal and valence. We map music pieces onto the arousal-valence plane, and classify the music emotion in that space. We extract 85 acoustic features from music signals, rank those by the information gain and choose the top k best features in the feature selection process. In order to map music pieces in the feature space onto the arousal-valence space, we apply genetic programming. The genetic programming is designed for finding an optimal formula which maps given music pieces to the arousal-valence space so that music emotions are effectively classified. k-NN and SVM methods which are widely used in classification are used for the classification of music emotions in the arousal-valence space. For verifying our method, we compare with other six existing methods on the same music data set. With this experiment, we confirm the proposed method is superior to others.     

Multi-level feature optimization and multimodal contextual fusion for sentiment analysis and emotion classification

The availability of the humongous amount of multimodal content on the internet, the multimodal sentiment classification, and emotion detection has become the most researched topic. The feature selection, context extraction, and multi-modal fusion are the most important challenges in multimodal sentiment classification and affective computing. To address these challenges this paper presents multilevel feature optimization and multimodal contextual fusion technique. The evolutionary computing based feature selection models extract a subset of features from multiple modalities. The contextual information between the neighboring utterances is extracted using bidirectional long-short-term-memory at multiple levels. Initially, bimodal fusion is performed by fusing a combination of two unimodal modalities at a time and finally, trimodal fusion is performed by fusing all three modalities. The result of the proposed method is demonstrated using two publically available datasets such as CMU-MOSI for sentiment classification and IEMOCAP for affective computing. Incorporating a subset of features and contextual information, the proposed model obtains better classification accuracy than the two standard baselines by over 3% and 6% in sentiment and emotion classification, respectively.     

An optimal and stable feature selection approach for traffic classification based on multi-criterion fusion

There is significant interest in the network management community about the need to identify the most optimal and stable features for network traffic data. In practice, feature selection techniques are used as a pre-processing step to eliminate meaningless features, and also as a tool to reveal the set of optimal features. Unfortunately, such techniques are often sensitive to a small variation in the traffic data. Thus, obtaining a stable feature set is crucial in enhancing the confidence of network operators. This paper proposes an robust approach, called the Global Optimization Approach (GOA), to identify both optimal and stable features, relying on multi-criterion fusion-based feature selection technique and an information-theoretic method. The proposed GOA first combines multiple well-known FS techniques to yield a possible optimal feature subsets across different traffic datasets; then the proposed adaptive threshold, which is based on entropy to extract the stable features. A new goodness measure is proposed within a Random Forest framework to estimate the final optimum feature subset. Experimental studies on network traffic data in spatial and temporal domains show that the proposed GOA approach outperforms the commonly used feature selection techniques for traffic classification task.     

Blog研究

Blog信息源和信息量迅速增长,并已通过频繁的链接和信息交互在互联网上构建了一个动态且紧密的社会网络,成为现实世界一个重要的信息来源.目前,Blog领域的研究主要集中在Blog的定义与识别、内容挖掘、社区发现、重要性分析、Blog搜索和作弊Blog识别等几个方面.大部分研究采用或借鉴了链接分析、自然语言处理等方面的技术和方法,也提出了一些针对Blog领域的特定方法.分析和比较了Blog领域的相关研究,并且讨论了研究中存在的问题,展望了未来的研究方向.     

基于字词融合特征的微博情绪识别方法

文本情绪识别是自然语言处理问题中的一项基本任务。该任务旨在通过分析文本判断该文本是否含有情绪。针对该任务,提出了一种基于字词融合特征的微博情绪识别方法。相对于传统方法,所提方法能够充分考虑微博语言的特点,充分利用字词融合特征提升识别性能。具体而言,首先将微博文本分别用字特征和词特征表示;然后利用LSTM模型(或双向LSTM模型)分别从字特征和词特征表示的微博文本中提取隐层特征;最后融合两组隐层特征,得到字词融合特征,从而进行情绪识别。实验结果表明,该方法能够获得更好的情绪识别性能。     

Knowledge discovery using neural approach for SME’s credit risk analysis problem in Turkey

This study proposes a knowledge discovery method that uses multilayer perceptron (MLP) based neural rule extraction (NRE) approach for credit risk analysis (CRA) of real-life small and medium enterprises (SMEs) in Turkey. A feature selection and extraction stage is followed by neural classification that produces accurate rule sets. In the first stage, the feature selection is achieved by decision tree (DT), recursive feature extraction with support vector machines (RFE-SVM) methods and the feature extraction is performed by factor analysis (FA), principal component analysis (PCA) methods. It is observed that the RFE-SVM approach gave the best result in terms of classification accuracy and minimal input dimension. Among various classifiers k-NN, MLP and SVM are compared in classification experiments. Then, the Continuous/Discrete Rule Extractor via Decision Tree Induction (CRED) algorithm is used to extract rules from the hidden units of a MLP for knowledge discovery. Here, the MLP makes a decision for customers as being “good” or “bad” and reveals the rules obtained at the final decision. In the experiments, Turkish SME database has 512 samples. The proposed approach validates the claim that is a viable alternative to other methods for knowledge discovery.     

基于本体特征的影评细粒度情感分类

针对中文影评情感分类中缺少特征属性及情感强度层面的粒度划分问题,提出一种基于本体特征的细粒度情感分类模型。首先,利用词频逆文档频率（TF-IDF）和TextRank算法提取电影特征,构建本体概念模型。其次,将电影特征属性和普鲁契克多维度情绪模型与双向长短时记忆网络（Bi-LSTM）融合,构建了在特征粒度层面和八分类情感强度下的细粒度情感分类模型。实验中,本体特征分析表明：观影人对故事属性关注度最高,继而是题材、人物、场景、导演等特征;模型性能分析表明：基于特征粒度和八分类情感强度,与应用情感词典、机器学习、Bi-LSTM网络算法在整体粒度和三分类情感强度层面的其他5个分类模型相比,该模型不仅有较高的F1值（0.93）,而且还能提供观影人对电影属性的情感偏好和情感强度参考,实现了中文影评更细粒度的情感分类。     

基于对抗式神经网络的多维度情绪回归

情绪分析是细粒度的情感分析任务,其目的是通过训练机器学习模型来判别文本中蕴含了何种情绪,是当前自然语言处理领域中的研究热点.情绪分析可细分为情绪分类与情绪回归两个任务.针对情绪回归任务,提出一种基于对抗式神经网络的多维度情绪回归方法.所提出的对抗式神经网络由3部分组成：特征抽取器、回归器、判别器.该方法旨在训练多个特征抽取器和回归器,以对输入文本的不同情绪维度进行打分.特征抽取器接受文本为输入,从文本中抽取针对不同情绪维度的特征;回归器接受由特征抽取器输出的特征为输入,对文本的不同情绪维度打分;判别器接受由特征抽取器输出的特征为输入,以判别输入的特征是针对何情绪维度.该方法借助判别器对不同的特征抽取器进行对抗式训练,从而获得能够抽取出泛化性更强的针对不同情绪维度的特征抽取器.在EMOBANK多维度情绪回归语料上的实验结果表明,该方法在EMOBANK新闻领域和小说领域的情绪回归上均取得了较为显著的性能提升,并在r值上超过了所有的基准系统,其中包括文本回归领域的先进系统.     

Fuzzy-rough feature selection aided support vector machines for Mars image classification

This paper presents a novel application of advanced machine learning techniques for Mars terrain image classification. Fuzzy-rough feature selection (FRFS) is adapted and then employed in conjunction with Support Vector Machines (SVMs) to construct image classifiers. These techniques are integrated to address problems in space engineering where the images are of many classes, large-scale, and diverse representational properties. The use of the adapted FRFS allows the induction of low-dimensionality feature sets from feature patterns of a much higher dimensionality. To evaluate the proposed work, K-Nearest Neighbours (KNNs) and decision trees (DTREEs) based image classifiers as well as information gain rank (IGR) based feature selection are also investigated here, as possible alternatives to the underlying machine learning techniques adopted. The results of systematic comparative studies demonstrate that in general, feature selection improves the performance of classifiers that are intended for use in high dimensional domains. In particular, the proposed approach helps to increase the classification accuracy, while enhancing classification efficiency by requiring considerably less features. This is evident in that the resultant SVM-based classifiers which utilise FRFS-selected features generally outperform KNN and DTREE based classifiers and those which use IGR-returned features. The work is therefore shown to be of great potential for on-board or ground-based image classification in future Mars rover missions.     

An optimal two stage feature selection for speech emotion recognition using acoustic features

Feature Fusion plays an important role in speech emotion recognition to improve the classification accuracy by combining the most popular acoustic features for speech emotion recognition like energy, pitch and mel frequency cepstral coefficients. However the performance of the system is not optimal because of the computational complexity of the system, which occurs due to high dimensional correlated feature set after feature fusion. In this paper, a two stage feature selection method is proposed. In first stage feature selection, appropriate features are selected and fused together for speech emotion recognition. In second stage feature selection, optimal feature subset selection techniques [sequential forward selection (SFS) and sequential floating forward selection (SFFS)] are used to eliminate the curse of dimensionality problem due to high dimensional feature vector after feature fusion. Finally the emotions are classified by using several classifiers like Linear Discriminant Analysis (LDA), Regularized Discriminant Analysis (RDA), Support Vector Machine (SVM) and K Nearest Neighbor (KNN). The performance of overall emotion recognition system is validated over Berlin and Spanish databases by considering classification rate. An optimal uncorrelated feature set is obtained by using SFS and SFFS individually. Results reveal that SFFS is a better choice as a feature subset selection method because SFS suffers from nesting problem i.e it is difficult to discard a feature after it is retained into the set. SFFS eliminates this nesting problem by making the set not to be fixed at any stage but floating up and down during the selection based on the objective function. Experimental results showed that the efficiency of the classifier is improved by 15–20 % with two stage feature selection method when compared with performance of the classifier with feature fusion.     

语音情感识别研究综述

针对语音情感识别研究体系进行综述。这一体系包括情感描述模型、情感语音数据库、特征提取与降维、情感分类与回归算法4个方面的内容。本文总结离散情感模型、维度情感模型和两模型间单向映射的情感描述方法；归纳出情感语音数据库选择的依据；细化了语音情感特征分类并列出了常用特征提取工具；最后对特征提取和情感分类与回归的常用算法特点进行凝练并总结深度学习研究进展，并提出情感语音识别领域需要解决的新问题、预测了发展趋势。     

Classification of emotions induced by music videos and correlation with participants’ rating

Emotional experience and preference play a vital role in selection of multimedia content for an individual. Brain electrical activity bears the emotional cues needed for emotion detection, but very modest research has been done to extract those cues. This paper presents a novel machine learning approach using Dual-Tree Complex Wavelet Packet Transform (DT-CWPT) time–frequency features from electroencephalogram (EEG) to detect emotions together with an analysis of brain activity in different emotional states. Firstly, DT-CWPT is used to extract time–frequency emotional features. Then non-redundant and most discriminating emotional features are selected through singular value decomposition (SVD), QR factorization with column pivoting (QRcp) and F-Ratio based feature selection (FS) method. The reduced emotional feature set is used to classify emotion using support vector machine (SVM) and validated by leave-one-out cross-validation scheme. Results confirm the robustness and consistency in classification of emotions from EEG signals and significant correlation between participants’ self assessed ratings with emotional features. It also gives an analysis of activities in brain region during different emotional states.     

表情和姿态的双模态情感识别

多模态情感识别是当前情感计算研究领域的重要内容,针对人脸表情和动作姿态开展双模态情感识别研究,提出一种基于双边稀疏偏最小二乘的表情和姿态的双模态情感识别方法.首先,从视频图像系列中分别提取表情和姿态两种模态的空时特征作为情感特征矢量.然后,通过双边稀疏偏最小二乘(BSPLS)的数据降维方法来进一步提取两组模态中的情感特征,并组合成新的情感特征向量.最后,采用了两种分类器来进行情感的分类识别.以国际上广泛采用的FABO表情和姿态的双模态情感数据库为实验数据,并与多种子空间方法(主成分分析、典型相关分析、偏最小二乘回归)进行对比实验来评估本文方法的识别性能.实验结果表明,两种模态融合后相比单模态更加有效,双边稀疏偏最小二乘(BSPLS)算法在几种方法中得到最高的情感识别率.     

基于DBN的多特征融合音乐情感分类方法

本文在音乐情感分类中的两个重要的环节：特征选择和分类器上进行了探索.在特征选择方面基于传统算法中单一特征无法全面表达音乐情感的问题,本文提出了多特征融合的方法,具体操作方式是用音色特征与韵律特征相结合作为音乐情感的符号表达;在分类器选择中,本文采用了在音频检索领域表现较好的深度置信网络进行音乐情感训练和分类.实验结果表明,该算法对音乐情感分类的表现较好,高于单一特征的分类方法和SVM分类的方法.     

维度语音情感识别研究综述

情感识别是多学科交叉的研究方向,涉及认知科学、心理学、信号处理、模式识别、人工智能等领域的研究热点,目的是使机器理解人类情感状态,进而实现自然人机交互.本文首先从心理学及认知学角度介绍了语音情感认知研究进展,详细介绍了情感的认知理论、维度理论、脑机制以及基于情感理论的计算模型,旨在为语音情感识别提供科学的情感理论模型.然后,从人工智能角度系统地总结了目前维度情感识别的研究现状和发展,包括语音维度情感数据库、特征提取、识别算法等技术要点.最后,分析了维度情感识别技术目前面临的挑战以及可能的解决思路,对未来研究方向进行了展望.