基于语义相似度及命名实体识别的主观题自动评分方法

人工智能技术目前已运用于诸多领域,其中在教育行业的一个重要应用为主观题评分系统。文本相似度计算是主观题评分的一大难点,目前采用的基于同义词词林的词语相似度计算方法已经了取得较好的效果,但文本过长会导致传统词语相似度计算方法性能下降。该文采用拓展的命名实体识别方法将主观题的候选答案中部分关键词提取出来,采用改进的同义词林词语相似度计算方法将候选关键词与主观题标准答案中目标关键词进行相似度计算。所提方法能有效提升词语匹配效率,在原同义词林词语相似度算法基础上,提升了性能,有效缩短了计算时间。     

基于HMM的语音信号情感识别研究

包含在语音信号中的情感信息是一种很重要的信息,它是人们感知事物不可缺少的部分。本文在语音识别的基础上提出了应用隐马尔可夫模型(HMM)进行语音信号情感识别的研究。从情感语音的分类、情感语音资料的获取、情感语音特征提取及情感语音识别等方面,讨论了应用连续隐马尔可夫模型进行情感识别的整个过程,并得到了比较理想的识别结果。     

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.     

Cross-task emotion recognition using eeg measures: first step towards practical application

Electroencephalographic (EEG)-based emotion recognition has received increasing attention in the field of human-computer interaction (HCI) recently, there however remains a number of challenges in building a generalized emotion recognition model, one of which includes the difficulty of an EEG-based emotion classifier trained on a specific task to handle other tasks. Little attention has been paid to this issue. The current study is to determine the feasibility of coping with this challenge using feature selection. 12 healthy volunteers were emotionally elicited when conducting picture induced and video induced tasks. Firstly, support vector machine (SVM) classifier was examined under within-task conditions (trained and tested on the same task) and cross-task conditions (trained on one task and tested on another task) for picture induced and video induced tasks. The within-task classification performed fairly well (classification accuracy: 51.6% for picture task and 94.4% for video task). Cross-task classification, however, deteriorated to low levels (around 44%). Trained and tested with the most robust feature subset selected by SVM-recursive feature elimination (RFE), the performance of cross-task classifier was significantly improved to above 68%. These results suggest that cross-task emotion recognition is feasible with proper methods and bring EEG-based emotion recognition models closer to being able to discriminate emotion states for any tasks.     

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。     

基于幅值滤波与分层特征融合策略的语音情感识别

针对语音情感识别在多语言联合数据集上识别准确率低的问题,提出了一种基于幅值滤波与分层特征融合策略的语 音情感识别方法。该方法首先对梅尔谱图内幅值分布规律进行幅值滤波,通过概率叠加扩大梅尔谱图内相近幅值之间的差 异,实现谱图内的高频强增益、低频弱增益;同时,通过概率相乘缩小梅尔谱图内相远幅值之间的差异,以显示谱图内中频的 细节部分。在此基础上,使用矩形卷积提取音频信号的时间动态特征,生成梅尔谱图动态特征图,并将其作为分层特征融合 策略的输入。分层特征融合策略通过压缩特征图来提取不同尺度的时间动态特征,并提取不同深度中的时间动态特征。在 多语言联合数据集 CER 上取得了84.44%的分类准确率。     

Power quality events recognition using a SVM-based method

In this paper, a novel SVM-based method for power quality event classification is proposed. A simple approach for feature extraction is introduced, based on the subtraction of the fundamental component from the acquired voltage signal. The resulting signal is presented to a support vector machine for event classification. Results from simulation are presented and compared with two other methods, the OTFR and the LCEC. The proposed method shown an improved performance followed by a reasonable computational cost.     

一种新的电能质量扰动信号压缩感知识别方法

针对现有电能质量扰动信号识别方法存在数据量大、准确率不高的不足,提出了一种基于压缩感知稀疏向量特征提取的电能质量扰动信号分类识别方法。该方法首先针对原始信号,利用压缩感知理论获取降维的测量信号,并基于?1范数正交匹配追踪算法获取稀疏向量。然后针对稀疏向量提取最大值、次大值、均方根、标准差、峭度和裕度因子等特征,作为神经网络的输入,实现电能质量扰动信号的分类识别。最后,针对六类典型电能质量扰动信号,开展仿真实验验证。仿真结果表明,现有识别方法需要处理的原始信号长度为1024,而所提方法特征提取时所处理的数据长度仅有30,从而大大减少了所需处理的数据量,并且由于实现了以非常少的数据量保存原有全部有用特征信息,因而更有利于提高识别准确率。通过与广泛采用的小波变换识别方法进行比较,所提方法的平均准确率高达98.71%,远远高于小波变换方法的92.86%。     

水利施工事故文本智能分析的BERT-BiLSTM混合模型

水利工程施工往往具有施工环境复杂、施工难度大的特点,施工事故频发。事故报告作为事故分析的文件,通常包含了事故发生的总结和原因,可以作为预防事故发生的依据。然而,目前水利领域的事故分析多依赖于现场管理人员的手工分析,不仅容易出错,而且耗时耗力。此外,现有的模型无法直接对水利事故文本进行高精度的分析和预测。因此,本文提出了一种结合变压器双向编码表示（BERT）和双向长短时记忆模型（BiLSTM）的混合深度学习模型深入分析水利工程施工事故原因。混合模型的上游采用BERT模型生成事故文本的字符级动态特征向量,模型下游基于改进的BiLSTM模型挖掘事故报告文本的语义特征,实现事故报告文本智能分析。最后,通过将本文提出的模型和目前先进的七种深度学习模型进行实验对比,对所提出的混合模型的有效性进行验证。结果表明,本文提出的混合模型优于其他几种深度学习算法,该模型可为水利施工事故的分析与决策提供算法支撑和依据。     

局部自适应复合绝缘子检测与憎水性分类网络

憎水性等级(Hydrophobicity Class,HC)是衡量绝缘子性能的重要指标之一。在实际环境的多种因素作用下绝缘子伞裙表面存在局部憎水性差异,为了准确识别绝缘子的性能,本文提出了一种基于深度学习的局部自适应绝缘子检测与憎水性分类模型。首先,通过绝缘子分割模块分离绝缘子与背景区域,为后续针对绝缘子区域的操作提供分割信息;然后将绝缘子区域划分为固定大小的图像块,在缩小分辨率减小运算难度的同时保留了绝缘子表面的细节信息;最后通过憎水性分类模块分析图像块内绝缘子的憎水性。实验使用巡检维护现场的绝缘子图片作为样本集,分阶段构建模型,分别对分割阶段和憎水性分类阶段的准确性进行评估。实验结果显示分割阶段模块能有效识别绝缘子和背景区域,交叉验证的测试集准确率均大于97.21%,并且憎水性分类阶段模块能准确分析绝缘子憎水性,对140幅测试图片的识别准确率达到98.65%。经过实验证明本文提出的模型在复杂自然环境中检测绝缘子性能是一种有效的解决方案。     

Cellular wave computer algorithms with spatial semantic embedding for handwritten text recognition

The recognition of cursive handwritten texts is a complex, in some cases unsolvable, task. One problem is that in most cases it is difficult or impossible to identify each letter, even if the words are separated. In our new method, the identification of letters is not needed due to the extensive and iterative use of semantic and morphological information of a given language. We are using a spatial feature code, generated by a cellular nonlinear network (CNN) based cellular wave computer algorithm, and combine it with the linguistic properties of the given language. Most general‐purpose handwriting recognition systems lack the ability to integrate linguistic background knowledge because they use it only for post‐processing recognition results. The high‐level a priori background knowledge is, however, crucial in human reading and similarly it can boost recognition rates dramatically in case of recognition systems. In our new system we do not treat the visual source as the only input: geometric and linguistic information are given equal importance. On the geometric side we use word‐level holistic feature detection without letter segmentation by analogic CNN algorithms designed for cellular wave computers (IEEE Trans. Circuits Syst. 1993; 40 :163–173; Cellular Neural Networks and Visual Computing, Foundations and Applications. Cambridge University Press: Cambridge, U.K., New York, 2002). The linguistic side is based on a morpho‐syntactic linguistic system (Proceedings of COLING‐2002, vol. II, Taipei, Taiwan, 2002; 1263–1267). A novel shape coding method is used to interface them, and their interaction is enhanced via an inverse filtering technique based on features that are global or of a low confidence value. A statistical context selection method is also applied to further reduce the output word lists. Copyright © 2008 John Wiley & Sons, Ltd.     

一种基于支持向量机的三维物体识别方法

提出从三维物体的二维图像中提取仿射不变傅氏描述子、色彩矩及纹理特征,组成一个25维的特征向量,送入支持向量机训练并用于三维物体识别。算法利用了仿射不变傅氏描述子在物体发生仿射形变时具有不变性,利用色彩矩和纹理特征区分形状相似但有不同颜色及纹理的物体,并引入支持向量机作为分类器。基于三维物体图像数据库COIL-100测试了算法的识别性能。当每个物体训练样本图像数量为36个（视角间隔10°）时达到了100％的识别率,进一步减少训练视角数量也达到较满意的识别性能。     

基于DTW算法的非侵入式家居负荷行为识别方法

随着电力负荷监测在生产和生活中的广泛应用,非侵入式负荷识别技术得到发展,具有很好的发展前景。针对非侵入式家用电器负荷的行为识别存在识别特征量难以优化,识别算法收敛性差等问题,提出了一种基于特征匹配度的家居负荷的行为识别方法。将家用负荷开、关时的暂态波形和功率变化值作为特征量,运用DTW算法计算测试模板与参考模板之间的相似度,有效的识别出家庭用电的各项负荷数据,并且通过三种案例来模拟负荷的识别过程,验证了算法的可行性。     

基于智能视觉物联网的油库人员行为识别与监测系统

针对油库区域人员行为监测响应慢、时效低的问题,提出一种基于智能视觉物联网的油库人员行为识别与监测系统。首先建立智能视觉物联网监测系统架构,满足信号采集、传输、处理与反馈需求;然后提出一种视频语义分析模型,将人体行为识别与人脸识别进行协同分析,实现对油库作业人员行为的分析与监测。经实验验证,在自建的数据库中,系统的人脸识别准确率达96.5%,人体行为识别准确率达85%,说明该系统可有效减少因人员的行为失误造成的油库安全事故,在油库的安全管理中具有很高的应用价值。     

地址模糊匹配在电力企业应用中的研究与实践

地址识别作为大数据应用中自然语言处理的重要场景之一,是重要且极具实用价值的技术手段。目前电力企业正在积极推进大数据应用演进进程,使用大数据技术为电力企业发展赋能的需求也愈加旺盛。而电力企业众多数据资产中的地址信息,作为联系设备—客户—企业的核心字段,其分析挖掘价值极高。首先,针对电力企业常见地址信息涉及的业务应用场景进行分析,瞄准需求要点,对核心应用场景的地址识别需求制定提取匹配方法;然后,基于数据样本对地址识别方法的准确性及运行速度进行研究分析,比对算法的实用性;最后,对算法进行了总结,并对未来算法的提升方式进行展望,以供电力企业相关业务在实际应用中进行参考。     

AGA-BP神经网络用于变压器超高频局部放电模式识别

结合自适应遗传算法(AGA)和BP算法各自的优点，本文构造了AGA—BP混合算法作为神经网络的学习算法。分别采用BP、AGA和AGA—BP神经网络对实验室中变压器超高频局部放电自动识别系统检测列的五种放电类型进行了模式识别。实验结果表明，AGA—BP神经网络既解决了BP神经网络对初始权值敏感和容易局部收敛的问题，又提高了AGA神经网络的收敛速度、稳定性和求解质量，具有较高的识别率和较强的推广能力。     

Fault localization in electrical power systems: A pattern recognition approach

Electrical power system is one of the most complex artificial systems in this world, which safe, steady, economical and reliable operation plays a very important part in social economic development, even in social stability. The fault in power system cannot be completely avoided. In this paper, we developed a method to resolve fault localization problems in power system. In our researches, based on real-time measurement of phasor measurement units, we used mainly pattern classification technology and linear discrimination principle of pattern recognition theory to search for laws of electrical quantity marked changes. The simulation results indicate that respectively study on the phase voltage, positive sequence voltage, negative sequence voltage, phase current, positive sequence current, negative sequence current of single-phase grounding faults and the positive sequence voltage, positive sequence current of three-phase short circuit faults, the pattern classification technology and linear discrimination principle are able to quickly and accurately identify the fault components and fault sections, and eventually accomplish fault isolation. In the study of electrical power systems, pattern recognition theory must have a good prospect of application.     

考虑信息动态表达的异常用电模式识别云边协同方法

异常用电识别是用电稽查、计量装置运行状态辨识的重要内容,对维护电网的安全运行和保障正常用户权益有重要的意义。已有方法为了识别用户的多元用电模式,在保证识别准确性的基础上容易造成计算过于复杂的问题,而考虑效率的简单计算方法又难以准确度量不同用电模式的相似性,因此难以兼顾计算效率与准确性;此外,将用电数据上传至云端集中计算会占用大量的网络带宽和计算资源,进一步限制了异常辨识的应用。为此,提出了一种考虑信息动态表达的异常用电模式识别云边协同方法。根据边缘端和云端的计算资源合理分配协作任务,实现了异常用电的云边协同识别。针对边缘服务器算力有限的问题,对用电数据进行动态压缩重表达,在缩减数据量的同时保证数据信息的准确性。云端在收到压缩数据后以分段加权动态时间规整距离作为压缩数据相似性度量的依据,基于自适应参数选择的密度聚类算法识别异常用电。基于实际数据集验证了所提方法的有效性。