Semisupervised learning of author‐specific emotions in micro‐blogs

Learning emotions from texts has been an active research topic in affective computing. However, the lack of reliable connection between emotions and language features has caused severely biased emotion predictions. Moreover, the author‐specific patterns in emotion expression could potentially affect emotion predictions, which has never been studied. In this paper, we propose a semisupervised learning algorithm to learn emotional features from large‐scaled micro‐blog documents with a Bayesian network, and introduce an emotion transition factor to generate the author‐specific emotion predictions. We infer the author‐specific emotions in micro‐blog streams through belief propagation, and learn the emotional features through an expectation maximization estimation procedure. We report results of single‐label and multilabel emotion predictions on a micro‐blog stream corpus, and analyze the improvements achieved by the semisupervised feature learning strategy and the incorporation of emotion transition patterns. Finally, we perform personality analysis based on the authors' emotion distribution and examine emotion distributions in the learned features. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Multi‐label emotion recognition of weblog sentence based on Bayesian networks

An increasing number of common users, in the Internet age, tend to express their emotions on the Web about everything they like or dislike. As a consequence, the number of all kinds of reviews, such as weblogs, production reviews, and news reviews, grows rapidly. This makes it difficult for people to understand the opinions of the reviews and obtain useful emotion information from such a huge number of reviews. Many scientists and researchers have attached more attention to emotion analysis of online information in the natural language processing field. Different from previous works, which just focused on the single‐label emotion analysis, this paper takes into account rich and delicate emotions and gives special regard to multi‐label emotion recognition for weblog sentences based on the Chinese emotion corpus (Ren‐CECps). Using the theory of Bayesian networks and probabilistic graphical model, the latent emotion variable and topic variable are employed to find out the complex emotions of weblog sentences. Our experimental results on the multi‐label emotion topic model demonstrate the effectiveness of the model in recognizing the polarity of sentence emotions. © 2015 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于情感词典与LDA模型的股市文本情感分析

建立了一种基于股票情感词典与LDA分析股票文本情感倾向的模型。针对股票文本情感分析中情感词典不全面与句子分析片面的问题,构建较为全面的股票情感词典,同时以句子的倾向性、程度性与相关性三方面分析股票文本情感。引入针对股票的词语、程度性词语与转折性词语构建较为全面的情感词典;抽取预处理之后的股票文本句子的情感词;利用句子算法计算句子倾向、程度向量,并对句子向量利用支持向量机(SVM)和K均值算法分类;利用LDA(latent dirichlet allocation)对情感词计算文档 主题、文档 词语概率分布,以此概率分布获取句子的相关性;综合句子的倾向性、程度性、相关性计算句子情感;最后,通过句子情感获取股票文本的情感倾向比例。通过对百度新闻经济板块收集的股票文本进行实验并与其他算法比较,该模型对句子与文本分类准确率提高到82.78%与84.14%。     

An Approach to Speech Emotion Classification Using k-NN and SVMs

The interaction between humans and machines has become an issue of concern in recent years. Besides facial ex-pressions or gestures, speech has been evidenced as one of the foremost promising modalities for automatic emotion recognition. Effective computing means to support HCI (Human-Computer Interaction) at a psychological level, al-lowing PCs to adjust their reactions as per human requirements. Therefore, the recognition of emotion is pivotal in High-level interactions. Each Emotion has distinctive properties that form us to recognize them. The acoustic signal produced for identical expression or sentence changes is essentially a direct result of biophysical changes, (for example, the stress instigated narrowing of the larynx) set off by emotions. This connection between acoustic cues and emotions made Speech Emotion Recognition one of the moving subjects of the emotive computing area. The most motivation behind a Speech Emotion Recognition algorithm is to observe the emotional condition of a speaker from recorded Speech signals. The results from the application of k-NN and OVA-SVM for MFCC features without and with a feature selection approach are presented in this research. The MFCC features from the audio signal were initially extracted to characterize the properties of emotional speech. Secondly, nine basic statistical measures were calculated from MFCC and 117-dimensional features were consequently obtained to train the classifiers for seven different classes (Anger, Happiness, Disgust, Fear, Sadness, Disgust, Boredom and Neutral) of emotions. Next, Classification was done in four steps. First, all the 117-features are classified using both classifiers. Second, the best classifier was found and then features were scaled to [-1, 1] and classified. In the third step, the with or without feature scaling which gives better performance was derived from the results of the second step and the classification was done for each of the basic sta-tistical measures separately. Finally, in the fourth step, the combination of statistical measures which gives better per-formance was derived using the forward feature selection method Experiments were carried out using k-NN with different k values and a linear OVA-based SVM classifier with different optimal values. Berlin emotional speech da-tabase for the German language was utilized for testing the planned methodology and recognition rates as high as 60% accomplished for the recognition of emotion from voice signal for the set of statistical measures (median, maximum, mean, Inter-quartile range, skewness). OVA-SVM performs better than k-NN and the use of the feature selection technique gives a high rate.     

基于遗传优化的多级SVM语音情感识别

针对语音情感识别中特征维数高、识别率较低的问题,提出利用遗传算法进行特征降维,并构建二叉树结构的多级支持向量机(SVM)分类器进行语音多类情感识别的方案.首先对语音信号预处理后提取常用的情感特征,由于涉及特征较多,存在数据的冗余,采用遗传算法对提取的特征进行优化筛选;然后使用选出的最具情感区分能力的特征训练二叉树结构的多级SVM分类模型.在包含7种情感的柏林情感语料库上进行实验,结果证明提出的语音情感识别方案的有效性.     

基于 MIC 心率变异性特征选择的情感识别研究

心率变异性分析能够在情感识别中发挥重要作用,为了建立心电与情感类别之间的精准模型,提出了基于最大信息系 数(maximal information coefficient,MIC)的特征选择方法。 使用 Aubt 数据库和设计情感诱发实验进行研究,首先提取了心率变 异性时域、频域、非线性及时频域 40 个特征参数,然后基于 MIC 方法结合支持向量机、随机森林、K 近邻算法进行情感建模。 结 果显示,基于 MIC 特征选择方法,使用 Aubt 数据库针对唤醒度、效价、4 类情感的分类准确度分别为 90%、89%、84%,并进一步 选用皮尔森相关系数、ANOVA 特征选择方法与 MIC 进行对比;诱发实验数据中的多种一对一情感识别率均高于 75%。 结果表 明基于 MIC 特征选择方法能够显著提高分类准确度,对基于心电信号进行情感识别具有重要意义。     

基于卷积神经网络的脑电信号情绪分类方法

情绪作为人脑的高级功能,对人们的心理健康和个性特征有很大的影响。通过对脑电情绪数据集进行情绪分类,能够为今后实时监控正常人或抑郁病人的情绪提供进一步理论及实践依据。因此文章运用公开的脑电情绪数据集所提取的微分熵特征,并使用传统的滑动平均和线性动态系统方法,采用深度学习中的卷积神经网络作为基本前提,设计了一个卷积神经网络的脑电信号情绪分类模型,其包括4个卷积层、4个最大池化层、2个全连接层和1个Softmax层,并采用批归一化使参数搜索问题变容易,抑制模型过拟合。实验结果表明,利用该模型对SEED数据集的3种情绪识别的平均准确率达到了98.73%,精确率、召回率和F1分数分别为99.69%、98.12%和98.86%,ROC曲线下面积达0.998。与最近的类似工作相比,该文提出的卷积神经网络结构对于脑电信号情绪分类具有一定优越性。     

基于SVM的汉语语音情感识别研究

随着信息技术的发展,对人机交互能力的要求不断提高,情感信息处理已成为提高人机交互能力的一个重要课题.本文提出了一种汉语语音情感分类方法,主要研究了4种基本的人类情感:高兴、愤怒、恐惧、悲伤.从汉语语音信号中提取了能量、基频、语速等特征,利用支持向量机方法识别,取得了43.7%的平均识别率.     

A hybrid approach for word emotion recognition

In recent years, with plenty of online resources constantly emerging, emotion recognition in text has become increasingly important in human–computer interaction. Word emotion plays a very important role in emotion analysis of sentences or documents. This paper proposes a hybrid approach to recognition of word emotion in the dimension of eight emotion categories with corresponding intensities based on the Chinese emotion corpus. First, we present a new algorithm of semantic similarity computation for aiding emotion intensity computation and design a new algorithm of emotion vector computation by making use of both morpheme characteristics and semantic relations. And then, we adopt support vector machine model for the secondary classification to the words whose emotions cannot be calculated by the semantic analysis algorithm. Our approach achieves the accuracy of 54.00% and 78.75% for exact match and all five types of hit, respectively, on the basis of the core emotion lexicon CL4. Experimental results show that the integration of morpheme characteristics and semantic relations can improve the classification accuracy efficiently. © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

On the simulation of fast settling charge pump PLLs up to fourth order

In this paper, we discuss three different models for the simulation of integer‐N charge‐pump phase‐locked loops (PLLs), namely the continuous‐time s‐domain and discrete time z‐domain approximations and the exact semi‐analytical time‐domain model. The limitations of the two approximated models are analyzed in terms of error in the computed settling time as a function of loop parameters, deriving practical conditions under which the different models are reliable for fast settling PLLs up to fourth order. Besides, output spectral purity analysis methods based upon the time‐domain model are introduced and the results are compared with those obtained by means of the s‐domain model in terms of phase noise and reference spur estimation. As a case study, we use the three models to analyze a fast switching PLL to be integrated in a frequency synthesizer for WiMedia MB‐OFDM UWB systems. Copyright © 2010 John Wiley & Sons, Ltd.     

A CCII‐based non‐inverting Schmitt trigger and its application as astable multivibrator for capacitive sensor interfacing

This paper presents a novel second‐generation current conveyor (CCII)‐based non‐inverting Schmitt trigger topology. By means of the use of only three resistances, it is possible to set easily the threshold values or, in addition, the trigger can be set also to work as a zero‐voltage comparator. The theoretical working principle has been confirmed through PSpice simulations implementing an integrated CCII, designed in a low‐cost standard complementary metal–oxide–semiconductor technology (Austria Micro Systems (AMS) 0.35 µm) with low‐voltage low‐power characteristics, and then by experimental tests on the fabricated printed circuit board prototype through the use of the commercial component AD844 (Analog Devices) as CCII. As its main application example, the presented trigger has been employed to implement an astable multivibrator proposed here as a capacitive sensor interface capable to accurately detect about five decades of capacitive variations in the range of [100 pF–5.5 μF] with a maximum relative error lower than ±10%. Copyright © 2016 John Wiley & Sons, Ltd.     

Crank‐shaped and Y‐shaped through‐hole interconnections filled with Au‐Sn solder

Crank‐shaped and Y‐shaped through‐hole interconnections (THI) filled with Au‐Sn solder are demonstrated in this study. They are expected to realize a higher‐density wafer‐level package (WLP) for electrical devices including the fields of MEMS (Micro Electro‐Mechanical Systems) and MOEMS (Micro Optical Electro‐Mechanical Systems) compared with conventional WLP using THI consisting of straight through‐holes. In this paper, formation techniques for crank‐shaped and Y‐shaped THI are described. In order to make crank‐shaped and Y‐shaped through‐holes in a substrate, both femtosecond laser irradiation and wet chemical etching are used. Crank‐shaped through‐holes with an 80‐μm opening and 400 μm deep have also been achieved. Gold (Au)‐tin (Sn) solder was filled into the holes using the molten metal suction method (MMSM). Airtightness of the THI was examined using the helium leakage test, and the estimated leakage rate was less than 1.0 × 10^?9 Pa·m³/s, which is sufficient for use in WLP applications. © 2012 Wiley Periodicals, Inc. Electr Eng Jpn, 179(1): 54–62, 2012; Published online in Wiley Online Library ( wileyonlinelibrary.com) . DOI 10.1002/eej.21134     

基于VMD-LSTMQR的滚动母线负荷区间预测

负荷区间预测能够对负荷出力变化进行概率化分析,准确量化不确定性因素对负荷的影响。相较于传统的点预测,区间预测更能直观反映负荷变化趋势,有助于保障电力系统的安全稳定运行。基于此,文中提出一种基于变分模态分解-长短期记忆神经网络分位数回归（VMD-LSTMQR）的滚动母线负荷区间预测方法。首先,文中采用VMD将原始母线负荷分解成一系列不同频率特征的子序列;接着,确定不同子序列的最优滚动步长并采用LSTMQR分别对不同子序列进行区间预测;最后,将不同子序列的区间预测进行重构,得到原始母线负荷预测结果。文中利用220 kV和10 kV母线负荷数据验证了所采用的区间预测模型相较于传统区间预测模型在预测精度、区间宽度方面得到明显改善。     

Integrated design of fault detection,isolation, and control for continuous‐time Markovian jump systems

In this paper, the problem of integrated fault detection, isolation, and control design of continuous‐time Markovian jump linear systems with uncertain transition probabilities is introduced and addressed for the first time in the literature. A single Markovian jump module designated as the integrated fault detection, isolation, and control under a mixed robust framework is considered to simultaneously achieve the desired detection, isolation, and control objectives. Conventional mixed robust approaches to the fault detection and isolation problem lead to conservative results due to the selection of identical Lyapunov matrices. Consequently, the extended linear matrix inequality methodology is utilized in this work to reduce the conservativeness of standard approaches by introducing additional matrix variables so that the coupling of Lyapunov matrices with the system matrices is eliminated. Simulation results for an application to the GE F‐404 aircraft engine system illustrate the effectiveness and capabilities of our proposed design methodologies. Comparisons with relevant work in the literature are also provided to demonstrate the advantages of our proposed solutions.     

Design of a qualitative classification model through fuzzy support vector machine with type‐2 fuzzy expected regression classifier preset

Methods of qualitative analysis, such as qualitative classification, have gained importance as an essential complement of existing quantitative analysis in numerous fields. Only a few models have been developed to deal with qualitative inputs in the form of type‐2 fuzzy(T2F) sets properly, given that traditional defuzzification method like the Karnik–Mendel algorithm performs dimensionality reduction at the cost of loss of information. To improve the situation, we define the expected value and variance of T2F set in this paper. By using a combination of them, we transfer the vertical three‐dimensional uncertainty of T2F set to horizontal range uncertainty without much distortion of information. Additionally, current classification models are unsuitable to the partial classification problem if an output is not fully assigned to a single class. We build a comprehensive qualitative classification model based on fuzzy support vector machine (FSVM) combined with type‐2 fuzzy expected regression (FER) to solve the partial classification problem as mentioned. This classifier (i.e. FER‐FSVM) makes it possible to achieve the discrimination of output while characterizing membership for each class in terms of multidimensional qualitative inputs (attributes) in the form of T2F sets. FER‐FSVM also can self‐learn the data structure and shift between FER or FSVM for classification automatically, thus largely improving the efficiency of the classification process. The new model is almost 7 times more efficient than FSVM, as shown by our empirical experiments. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

TFSM‐based dialogue management model framework for affective dialogue systems

A new dialogue management model for affective dialogue system, which aims to provide a service of information inquiry and affective interaction, is proposed in this paper. First, we construct two finite state machines (TFSM) to model the user and the system, respectively, and simulate the dialogue process as an information exchange between the two state machines. All possible state transitions in dialogue and its probabilities of the user are summarized as a user model, which is helpful for the system to inference and predict the user's internal states. Second, we further discuss the implementation methods of information inquiry and emotional response modules. Finally, we employ the return function of partially observable Markov decision processes (POMDP) model to analyze and evaluate the TFSM‐based dialogue management model. The experimental results not only show the relationships between the average returns, recognition error rates, and state transition probabilities but also confirm that our TFSM‐based dialogue management model outperforms the conventional FSM model. © 2015 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Survey of integrated‐circuit‐oscillator phase‐noise analysis

This tutorial distills the salient phase‐noise analysis concepts and key equations developed over the last 75 years relevant to integrated circuit oscillators. Oscillator phase and amplitude fluctuations have been studied since at least 1938 when Berstein solved the Fokker–Planck equations for the phase/amplitude distributions of a resonant oscillator. The principal contribution of this work is the organized, unified presentation of eclectic phase‐noise analysis techniques, facilitating their application to integrated circuit oscillator design. Furthermore, we demonstrate that all these methods boil down to obtaining three things: (1) noise modulation function; (2) noise transfer function; and (3) current‐controlled oscillator gain. For each method, this paper provides a short background explanation of the technique, a step‐by‐step procedure of how to apply the method to hand calculation/computer simulation, and a worked example to demonstrate how to analyze a practical oscillator circuit with that method. This survey article chiefly deals with phase‐noise analysis methods, so to restrict its scope, we limit our discussion to the following: (1) analyzing integrated circuit metal–oxide–semiconductor/bipolar junction transistor‐based LC, delay, and ring oscillator topologies; (2) considering a few oscillator harmonics in our analysis; (3) analyzing thermal/flicker intrinsic device‐noise sources rather than environmental/parametric noise/wander; (4) providing mainly qualitative amplitude‐noise discussions; and (5) omitting measurement methods/phase‐noise reduction techniques. Copyright © 2013 John Wiley & Sons, Ltd.     

Stability and stabilization in switched discrete‐time systems

In this paper, we investigate the stability and stabilization problem for discrete‐time switched systems. We consider a probabilistic case where the system is switched among different subsystems, and the probability of each subsystem being active is defined as its occurrence probability. The relationship between the developed model of the switched system and the Markovian jump system is analyzed. For a switched system with a known subsystem occurrence probabilities, we give a stochastic stability criterion in terms of a linear matrix inequality. Then, we extend the results to a more practical case where the subsystem occurrence probabilities of switching are known to be constant, but their specific values are only known with some uncertainty. A new iterative approach is employed to choose the switching law between the subsystems. For unstable switched systems, mode‐dependent state feedback and static output feedback controllers are developed to achieve the stabilization objective. Finally, several simulation examples are presented to show the efficacy of the proposed criteria and methods. Copyright © 2012 John Wiley & Sons, Ltd.     

Analysis of lip motion change arising as a result of amusement feeling

This letter develops a novel algorithm for lip motion analysis arising as a result of feeling amused after watching television comedy programs as emotion‐eliciting stimuli. The analysis is based on measuring the variance of new lip motion features extracted from pronouncing some sentences after watching the stimuli. The results of the conducted experiment show that these features can serve as efficient indices for determining the occurrence of amusement feelings. © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Improving Parsing of ‘BA’ Sentences for Machine Translation

The research on Chinese‐Japanese machine translation has been lasting for many years, and now this research field is increasingly thoroughly refined. In practical machine translation system, the processing of a simple and short Chinese sentence has somewhat good results. However, the translation of complex long Chinese sentence still has difficulties. For example, these systems are still unable to solve the translation problem of complex ‘BA’ sentences. In this article a new method of parsing of ‘BA’ sentence for machine translation based on valency theory is proposed. A ‘BA’ sentence is one that has a prepositional word ‘BA’. The structural character of a ‘BA’ sentence is that the original verb is behind the object word. The object word after the ‘BA’ preposition is used as an adverbial modifier of an active word. First, a large number of grammar items from Chinese grammar books are collected, and some elementary judgment rules are set by classifying and including the collected grammar items. Then, these judgment rules are put into use in actual Chinese language and are modified by checking their results instantly. Rules are checked and modified by using the statistical information from an actual corpus. Then, a five‐segment model used for ‘BA’ sentence translation is brought forward after the above mentioned analysis. Finally, we applied this proposed model into our developed machine translation system and evaluated the experimental results. It achieved a 91.3% rate of accuracy and the satisfying result verified effectiveness of our five‐segment model for ‘BA’ sentence translation. Copyright © 2007 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.