检索结果-南通市图书馆

在线全文

学校读者我要写书评

暂无评论

基于字句动态特征和自注意力的情感分析方法

数据采集与处理 2024年第1期39卷 193-203页

作者：刘强朱金森赵龙龙沙宇晨刘尚东季一木南京邮电大学计算机学院南京210023 南京邮电大学高性能计算与大数据处理研究所南京210023

针对评论情感分析任务中文本长度失衡引起的特征稀疏、特征缺失和提取信息不全等问题,提出了一种基于字句动态特征和自注意力的情感分析方法。首先基于预训练模型对评论进行动态特征编码,使用句向量对不足固定长度的部分进行补全,并表... 详细信息

针对评论情感分析任务中文本长度失衡引起的特征稀疏、特征缺失和提取信息不全等问题,提出了一种基于字句动态特征和自注意力的情感分析方法。首先基于预训练模型对评论进行动态特征编码,使用句向量对不足固定长度的部分进行补全,并表征超出的截断部分,以缓解批训练下文本尺寸失衡引起的特征稀疏和特征缺失问题。然后使用基于自注意力机制的特征重组方法动态整合字句融合特征,并优化权重参数以降低计算和训练时间复杂度。最后在开源数据集上分别进行了消融实验和对比实验。测试结果表明,本文方法在准确率上有较优的改进效果。

关键词：情感分析特征编码预训练模型自注意力机制权重参数

Collaborative filtering recommendation algorithm based on interactive data classification

在线全文

学校读者我要写书评

暂无评论

The Journal of China Universities of Posts and Telecommunications 2020年第5期27卷 1-12页

作者： Ji Yimu Li Ke liu shangdong liu Qiang Yao Haichang Li Kui School of Computer Science Nanjing University of Posts and TelecommunicationsNanjing 210023China Jiangsu High Technology Research Key Laboratory for Wireless Sensor Networks Nanjing University of Posts and TelocommunicutionsNanjing 210003China Institue of High Performance Computing and Bigdata Nanjing University of Posts and TelecommunicationsNanjing 210003China Nanjing Center of HPC China Nanjing University of Posts and TelocommunicutionsNanjing 210003China Jiangsu HPC and Intelligent Processing Engineer Research Center Nanjing University of Posts and Telocommunicutions Nanjing 210003China

In the matrix factorization(MF)based collaborative filtering recommendation method,the most critical part is to deal with the interaction between the features of users and *** mainstream approach is to use the inner product for MF to describe the user-item ***,as a shallow model,MF has its limitations in describing the relationship between *** addition,when the size of the data is large,the performance of MF is often poor due to data sparsity and *** paper presents a model called PIDC,short for potential interaction data clustering based deep learning ***,it uses classifiers to filter and cluster recommended items to solve the problem of sparse training ***,it combines MF and multi-layer perceptron(MLP)to optimize the prediction effect,and the limitation of inner product on the model expression ability is *** proposed model PIDC is tested on two *** experimental results show that compared with the existing benchmark algorithm,the model improved the recommendation effect.

关键词： personalized recommendation deep learning clustering collaborative filtering

Research on Security of Key Algorithms in Intelligent Driving System

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2019年第1期28卷 29-38页

作者： liu shangdong WU Ye JI Yimu CHEN Chen BI Qiang JIAO Zhipeng GONG Jian WANG Ruchuan School of Computer Science and Engineering Southeast University School of Computer Science Nanjing University of Posts and Telecommunications Jiangsu Provincial Key Laboratory of Computer Network Technology Jiangsu High Technology Research Key Laboratory for Wireless Sensor Networks

With the rapid development of the smart driving technology, the security of smart driving algorithms is becoming more and more important. Four core smart driving algorithms are determined by studying the architecture of smart driving algorithm. These algorithms comprise local path planning, pedestrian detection, lane detection and obstacle detection. The security issues of these algorithms are investigated by closely examining the work carried out by the algorithms. We found that there are vulnerabilities in all four algorithms. These vulnerabilities can cause abnormality and even road accidents for the smart cars. The final experiment shows that the vulnerabilities of these algorithms do exist under certain circumstances and therefore have high security risks. This study will lay a foundation to improve the security of the smart driving system.

关键词： Smart driving algorithm Local path planning Pedestrian detection Lane line detection Obstacle detection

LBA-EC: Load Balancing Algorithm Based on Weighted Bipartite Graph for Edge Computing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2023年第2期32卷 313-324页

作者： SHAO Sisi liu shangdong LI Kui YOU Shuai QIU Huajie YAO Xiaoliang JI Yimu School of Computer Science Nanjing University of Posts and Telecommunications Institue of High Performance Computing and Bigdata Nanjing University of Posts and Telecommunications Nanjing Center of HPC China Jiangsu HPC and Intelligent Processing Engineer Research Center

Compared with cloud computing environment, edge computing has many choices of service providers due to different deployment environments. The flexibility of edge computing makes the environment more complex. The current edge computing architecture has the problems of scattered computing resources and limited resources of single computing node. When the edge node carries too many task requests, the makespan of the task will be delayed. We propose a load balancing algorithm based on weighted bipartite graph for edge computing(LBA-EC), which makes full use of network edge resources, reduces user delay, and improves user service experience. The algorithm is divided into two phases for task scheduling. In the first phase, the tasks are matched to different edge servers. In the second phase, the tasks are optimally allocated to different containers in the edge server to execute according to the two indicators of energy consumption and completion time. The simulations and experimental results show that our algorithm can effectively map all tasks to available resources with a shorter completion time.

关键词： Energy consumption Processor scheduling Load management Scheduling Bipartite graph Delays Servers

A DHR executor selection algorithm based on historical credibility and dissimilarity clustering

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2023年第11期66卷 288-301页

作者： Sisi SHAO Yimu JI Weili ZHANG shangdong liu Fei JIANG Zhigang CAO Fei WU Fukang ZENG Jun ZUO Longfei ZHOU School of Computer Science Nanjing University of Posts and Telecommunications Institue of High Performance Computing and Bigdata Nanjing University of Posts and Telecommunications Purple Mountain Laboratories Nanjing Center of HPC China Jiangsu HPC and Intelligent Processing Engineer Research Center School of Foreign Languages Information Engineering University The 28th Research Institute of China Electronics Technology Group Corporation

The security of the dynamic heterogeneous redundancy(DHR) architecture relies on the heterogeneity of its executors, which also defines the vulnerability of the mimic system. In order to select executors with reliable and significant dissimilarity as service executors, we propose a DHR executor selection algorithm based on historical credibility and dissimilarity clustering(HCDC), which adds two metrics of executor historical credibility and dissimilarity. First, to maximize the difference between heterogeneous executor pools,clustering is performed based on the dissimilarity of the executor. Second, the executor with the highest historical credibility is selected from the heterogeneous executor pool as the candidate pool. The historical credibility is dynamically updated by the negative feedback control based on the results of the multi-mode adjudicator. Finally, the dynamic scheduling algorithm selects the executors from the candidate pool to form the set of service executors. The simulation results demonstrate that, in comparison to existing methods,the algorithm reduces the attack success rate and average failure rate while increasing system reliability.

关键词： endogenous security mimic defense DHR executor selection historical credibility

Ferroelectric-controlled graphene plasmonic surfaces for all-optical neuromorphic vision

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第3期67卷 765-773页

作者： CHEN JianBo liu Yu LI shangdong LIN Lin LI YaDong HUANG Wen GUO JunXiong School of Electronic Information and Electrical Engineering Institute for Advanced StudyChengdu UniversityChengdu 610106China School of Integrated Circuits Tsinghua UniversityBeijing 100084China School of Integrated Circuits(National Exemplary Schoolof Microelectronics) Universityof Electronic Science and Technologyof China Chengdu 610054China Jincheng Research Institute of Opto-mechatronics Industry Jincheng 048000China Shanxi Key Laboratory of Advanced Semiconductor Optoelectronic Devices and Integrated Systems Jincheng 048000China Engineering Research Center of Digital Imaging and Display Ministry of EducationSoochow UniversitySuzhou 215006China

Artificial visual systems can recognize desired objects and information from complex environments, and are therefore highly desired for pattern recognition, object detection, and imaging applications. However, state-of-the-art artificial visual systems with high recognition performances that typically consist of electronic devices face the challenges of requiring huge storage space and high power consumption owing to redundant data. Here, we report a terahertz(THz) frequency-selective surface using a graphene split-ring resonator driven by ferroelectric polarization for efficient visual system applications. The downward polarization of the ferroelectric material offers an ultrahigh electrostatic field for doping p-type graphene with an anticipated Fermi level. By optimizing the geometric parameters of the devices and modulating the carrier behaviors of graphene, our plasmonic devices exhibit a tunable spectral response in a range of 1.7–6.0 THz with continuous transmission values. The alloptical neural network using graphene plasmonic surfaces designed in this study exhibited excellent performance in visual preprocessing and convolutional filtering and achieved an ultrahigh recognition accuracy of up to 99.3% in training the Modified National Institute of Standards and Technology(MNIST) handwritten digit dataset. These features demonstrate the great potential of graphene plasmonic devices for future smart artificial vision systems.

关键词： graphene plasmon ferroelectric frequency selective surface artificial vision system

基于国产处理器的智能大数据一体机架构及应用研究

在线全文

学校读者我要写书评

暂无评论

南京邮电大学学报（自然科学版） 2024年第4期44卷 1-16页

作者：陈国良汤晓宇尤帅姚小良梅超君林时俊刘尚东吴少刚孙雅薇王汝传季一木南京邮电大学计算机学院江苏南京210023 南京邮电大学高性能计算与大数据处理研究所江苏南京210023 国家高性能计算中心南京分中心江苏南京210023 龙芯中科技术股份有限公司北京101111 北方信息控制研究院集团有限公司江苏南京211153

分析比较国内外智能芯片产品现状,概述智能芯片的相关软硬件技术栈,针对国内大数据与人工智能产业支撑体系尚不完善、尚未建立国产化信息处理安全保障体系的问题,提出基于国产芯片自主研发的智能大数据一体机研究与设计。解决基础设施... 详细信息

分析比较国内外智能芯片产品现状,概述智能芯片的相关软硬件技术栈,针对国内大数据与人工智能产业支撑体系尚不完善、尚未建立国产化信息处理安全保障体系的问题,提出基于国产芯片自主研发的智能大数据一体机研究与设计。解决基础设施的运维部署、海量数据存储以及时效性处理等问题,通过对比异构计算系统的分析研究,文中提出的基于国产CPU和NPU异构硬件加速的一体化计算平台具有信息安全、领域普适性强、运行效率高等特点。

关键词：国产芯片大数据一体机大数据处理智能计算信息安全

同方期刊数据库

THE SURFACE REACTION ON LIQUID-SOLID INTERFACE OF“AMINO ACID-Cu（II）-MnO2”SYSTEM IN SEA WATER

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Oceanology and Limnology 1988年第3期6卷 258-271页

作者：王修林张正斌刘莲生 shangdong College of Oceanology

The results obtained from the Cu（Ⅱ）%-pH curve of the system in the presence of added amino acid indicated that the added amino acids of higher concentration can inhibit the Cu（Ⅱ）-δ-MnO2 ion exchange and that their inhibiting effects decrease in the following sequence when the initial amino acid concentrations are 20.00 ppm: histidine > cystine > glutamic acid > proline > alanine > glycine. The determination of amino acid in solution performed with an amino acid analyser showed that added amino acids react chemically on the surface of the δ-MnO2 According to thermodynamic equilibria relations in the system, amino acids can be changed to amines because of their decarboxylation on the surface of the δ-MnO2. Considering （1） the effect-inhibiting factor of amine F （β, Ka, pH） representing the complex potential of amine with Cu（Ⅱ） insolution and （2） the acidic constant of - NH3+ group in the amine molecule representing the ability of cation exchange of the amine with δ-MnO2, the inhibiting seqence of

关键词：占一 THE SURFACE REACTION ON LIQUID-SOLID INTERFACE OF AMINO ACID-Cu II SYSTEM IN SEA WATER ppm

中美两国中小学体育教学中教师指导方式之比较

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

体育科学 1993年第2期 21-23页

作者：刘卓山东师范大学济南 250014

运用社会调查、文献分析和比较等方法,对中美两国中小学体育教学中的“教师指导方式”在教学目标、教学程序和课的结构三个方面的表现形式和特点进行了比较和分析,并得出如下结论:1.中美两国“教师指导方式”因分别受“形为主义”和“... 详细信息

运用社会调查、文献分析和比较等方法,对中美两国中小学体育教学中的“教师指导方式”在教学目标、教学程序和课的结构三个方面的表现形式和特点进行了比较和分析,并得出如下结论:1.中美两国“教师指导方式”因分别受“形为主义”和“人本主义”教育思想的影响而产生差异;2.影响两国“教师指导方式”的上述两种思想本身有共同之不足;3.两种“教师指导方式”对培养“全面发展的人来讲,均有力所不及之处。

关键词：教学指导方式垂直式教学模式水平式教学模式局部教学情境整体社会环境