检索结果-内蒙古大学图书馆

International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery

作者： Wuying Liu Lin Wang Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies

ISBN: (纸本)9781509040940

With the fast-paced prevalence of smartphones, binary short text classification (STC) is becoming a basic and challenging issue, and relevant STC algorithms can be successfully used in spam filtering for short message service (SMS), wechat, microblogging, and so on. In this manuscript, we address the structural feature of SMS documents and propose a structural learning framework, which decomposes the complex binary STC problem according to the SMS document structure, and predicts the final category by combining several subpredictions. Supported by our index of string-frequence, we also implement some STC domain classifiers. The experimental results show that the performance of two previous STC algorithms can be upgraded by the structural learning framework, and our STC domain classifiers can achieve the state-of-the-art performance on the task of Chinese SMS spam filtering within the structural learning framework.

关键词： Structural Learning Framework Binary Short Text Classification Index of String-Frequence SMS Spam Filtering TREC Spam Evaluation

来源：评论

学校读者我要写书评

暂无评论

An Effective RAID Data Layout for Object-Based De-duplication Backup System

引用

Chinese Journal of Electronics 2016年第5期25卷 832-840页

作者： YAN Fang TAN Yu'an ZHANG Quanxin WU Fei CHENG Zijing ZHENG Jun School of Computer Science and Technology Beijing Institute of Technology School of Information Beijing Wuzi University Beijing Engineering Research Center of Massive Language Information Processing and Cloud Computing Application Wuhan National Laboratory for Optoelectronics Huazhong University of Science and Technology Space Star Technology CO. LtdChina Academy of Space Technology

We propose an alternative Redundant array of independent disks(RAID) data layout, Asymmetrical grouping data organization(AGDO), for object-based data de-duplication backup system. Object-based data deduplication is an effective solution for detecting duplicate data for compound files. We designed an asymmetrical grouping strategy, the disk in the array are partitioned into different groups, and in each group parallelism data access scheme is adopted, different types of objects are stored in different groups with maybe different group size, it can set and manage dynamic group size by using dynamic disk group adjustment algorithms. The performance of AGDO is evaluated and proved to be sufficient for the continuous storage application. The result is that disk accesses are concentrated in a part of the disks over a long time period and reduces the power consumption to 25% in a 10-disk configuration. Moreover, object-based de-duplication combined with AGDO has great potential in increasing data restoration speed for compound files. We have shown that this combination makes average restoration speed improved 11%.

关键词： Data layout Object-based deduplication Backup Energy aware Redundant array of independent disks(RAID)

来源：评论

学校读者我要写书评

暂无评论

Mismatched training data enhancement for automatic recognition of children's speech using DNN-HMM 10th

Mismatched training data enhancement for automatic recogniti...

引用

10th International Symposium on Chinese Spoken language Processing, ISCSLP 2016

作者： Qian, Mengjie McLaughlin, Ian Quo, Wu Dai, Lirong National Engineering Laboratory of Speech and Language Information Processing University of Science and Technology of China Hefei Anhui China School of Computing University of Kent Medway Kent United Kingdom

ISBN: (纸本)9781509042937

The increasing profusion of commercial automatic speech recognition technology applications has been driven by big-data techniques, using high quality labelled speech datasets. Children's speech has greater time and frequency domain variability than typical adult speech, lacks good large scale training data, and presents difficulties relating to capture quality. Each of these factors reduces the performance of systems that automatically recognise children's speech. In this paper, children's speech recognition is investigated using a hybrid acoustic modelling approach based on deep neural networks and Gaussian mixture models with hidden Markov model back ends. We explore the incorporation of mismatched training data to achieve a better acoustic model and improve performance in the face of limited training data, as well as training data augmentation using noise. We also explore two arrangements for vocal tract length normalisation and a gender-based data selection technique suitable for training a children's speech recogniser. © 2016 IEEE.

关键词： Hidden Markov models

来源：评论

学校读者我要写书评

暂无评论

Compact convolutional neural network transfer learning for small-scale image classification

Compact convolutional neural network transfer learning for s...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zengxi Li Yan Song Ian Mcloughlin Lirong Dai USTC National Engineering Laboratory of Speech and Language Information Processing School of Computing University of Kent

ISBN: (纸本)9781479999897

Transfer learning methods have demonstrated state-of-the-art performance on various small-scale image classification tasks. This is generally achieved by exploiting the information from an ImageNet convolution neural network (ImageNet CNN). However, the transferred CNN model is generally with high computational complexity and storage requirement. It raises the issue for real-world applications, especially for some portable devices like phones and tablets without high-performance GPUs. Several approximation methods have been proposed to reduce the complexity by reconstructing the linear or non-linear filters (responses) in convolutional layers with a series of small ones., In this paper, we present a compact CNN transfer learning method for small-scale image classification. Specifically, it can be decomposed into fine-tuning and joint learning stages. In fine-tuning stage, a high-performance target CNN is trained by transferring information from the ImageNet CNN. In joint learning stage, a compact target CNN is optimized based on ground-truth labels, jointly with the predictions of the high-performance target CNN. The experimental results on CIFAR-10 and MIT Indoor Scene demonstrate the effectiveness and efficiency of our proposed method.

关键词： Training Learning systems Computational modeling Predictive models Computational complexity Linear programming

来源：评论

学校读者我要写书评

暂无评论

A Note on Determine the Greatest Common Subfamily of Two NFSRs by Grbner Basis

引用

Journal of Systems Science & Complexity 2015年第5期28卷 1231-1242页

作者： WANG Zhongxiao QI Wenfeng TIAN Tian Luoyang University of Foreign Language State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou Information Science and Technology Institute

For nonlinear feedback shift registers(NFSRs), their greatest common subfamily may be not unique. Given two NFSRs, the authors only consider the case that their greatest common subfamily exists and is unique. If the greatest common subfamily is exactly the set of all sequences which can be generated by both of them, the authors can determine it by Grbner basis theory. Otherwise, the authors can determine it under some conditions and partly solve the problem.

关键词： Greatest common subfamily Grbner basis nonlinear feedback shift register stream cipher

来源：评论

学校读者我要写书评

暂无评论

Natural Logic Inference for Emotion Detection

Natural Logic Inference for Emotion Detection

引用

第十六届全国计算语言学学术会议暨第五届基于自然标注大数据的自然语言处理国际学术研讨会

作者： Han Ren Wenhe Feng Xia Li Yafeng Ren Maofu Liu Laboratory of Language Engineering and Computing Guangdong University of Foreign StudiesGuangzhou 5 Collaborative Innovation Center for Language Research and Services Guangdong University of Foreign S College of Computer Science and Technology Wuhan University of Science and TechnologyWuhanHubei 43

ISBN: (纸本)9783319690049

Current research on emotion detection focuses on the recognizingexplicit emotion expressions in *** this paper,we propose an approach based on textual inference to detect implicit emotion expressions,that is,to cap-ture emotion detection as an logical inference *** approach builds a natu-ral logic system,in which emotional detection are decomposed into a series of logical inference *** system also employ inference knowledge from textural inference resources for reasoning complex expressions in emotional *** results show the efficiency in detecting implicit emotionalexpressions.

关键词： Natural Logic Textual Inference Emotion Detection Implicit Emotional Expression

来源：评论

学校读者我要写书评

暂无评论

Graphical representations of data improve student understanding of measurement and uncertainty: An eye-tracking study

引用

Physical Review Physics Education Research 2017年第2期13卷 020125-020125页

作者： Ana Susac Andreja Bubic Petra Martinjak Maja Planinic Marijan Palmovic Department of Physics Faculty of Science University of Zagreb Bijenicka 32 10000 Zagreb Croatia Department of Applied Physics Faculty of Electrical Engineering and Computing University of Zagreb Unska 3 10000 Zagreb Croatia Chair for Psychology Faculty of Humanities and Social Sciences University of Split Sinjska 2 21000 Split Croatia Laboratory for Psycholinguistic Research Department of Speech and Language Pathology University of Zagreb Borongajska cesta 83h 10000 Zagreb Croatia

Developing a better understanding of the measurement process and measurement uncertainty is one of the main goals of university physics laboratory courses. This study investigated the influence of graphical representation of data on student understanding and interpreting of measurement results. A sample of 101 undergraduate students (48 first year students and 53 third and fifth year students) from the Department of Physics, University of Zagreb were tested with a paper-and-pencil test consisting of eight multiple-choice test items about measurement uncertainties. One version of the test items included graphical representations of the measurement data. About half of the students solved that version of the test while the remaining students solved the same test without graphical representations. The results have shown that the students who had the graphical representation of data scored higher than their colleagues without graphical representation. In the second part of the study, measurements of eye movements were carried out on a sample of thirty undergraduate students from the Department of Physics, University of Zagreb while students were solving the same test on a computer screen. The results revealed that students who had the graphical representation of data spent considerably less time viewing the numerical data than the other group of students. These results indicate that graphical representation may be beneficial for data processing and data comparison. Graphical representation helps with visualization of data and therefore reduces the cognitive load on students while performing measurement data analysis, so students should be encouraged to use it.

关键词： Concepts & principles Instructional strategies Lower undergraduate students

来源：评论

学校读者我要写书评

暂无评论

BLEUS-syn:Cilin-based Smoothed BLEU

BLEUS-syn:Cilin-based Smoothed BLEU

引用

第十二届全国机器翻译研讨会

作者： Junting Yu Wuying Liu Hongye He Mianzhu Yi Luoyang University of Foreign Languages Luoyang 471003 China Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies Guangzhou 5

Machine Translation(MT)evaluation is very important for a MT *** this paper,we investigate an improved Cilin-based smoothed BLEU(BLEUS-syn).As the possible cases that the short translation or English abbreviations in candidate may cause unigram have no matches,this evaluation metric smoothed the traditional BLEUS *** applied synonym substitution in unigram matching,and calculated the other 2~*** performed experiments in Russian and Chinese bilingual sentence data set and evaluated the output translations of online translation systems such as Google,Baidu,Bing and *** experimental results show that the effectiveness of our BLEUS-syn and traditional BLEUS are *** performance of Baidu is the best,that of Youdao is the second,and that of Bing is the *** BLEUS-syn can greatly enhance the performance of traditional BLEUS *** makes the Baidu BLEUS value improve 6.81 percent,Youdao improve 6.98 percent,Google 7.82 percent,and Bing 7.55 percent.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Lexical-Index-based Japanese Syntax Matching

Lexical-Index-based Japanese Syntax Matching

引用

International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery

作者： Wuying Liu Lin Wang Xing Zhang Laboratory of Language Engineering and Computing Guangdong University of Foreign Studies Guangzhou 510420 China Luoyang University of Foreign Languages Luoyang 471003 China

ISBN: (纸本)9781509040940

Syntax matching is a challenging basic issue, and related algorithms can be widely used in natural language processing. This paper addresses the problem of how to efficiently match a sentence with the most similar syntactic structure to a given Japanese sentence from a big set of Japanese sentences, designs a novel lexical index data structure of hiratoken-sentence index (HSI) according to our Japanese syntax identification hypothesis of hiragana token, and proposes a HSI-based Japanese syntax matching (HSIJSM) algorithm. Supported by the HSI data structure, the HSIJSM algorithm can approximately get the syntactic similarity after the fast calculating of formal similarity between two Japanese sentences. The experimental results show that the HSIJSM algorithm can achieve the preferable performance with greatly reduced time costs.

关键词： Hiragana Token Japanese Syntax Matching Hiratoken-Sentence Index Syntactic Similarity Formal Similarity

来源：评论

学校读者我要写书评

暂无评论

Enhanced null message algorithm for hybrid parallel simulation systems with large disparity in time step

Enhanced null message algorithm for hybrid parallel simulati...

引用

International Workshop/Symposium on Distributed Simulation and Real-Time Applications (DS-RT)

作者： Bin Wang Yanlong Zhai Zishuo Wang Han Zhang Duzheng Qing Beijing Engineering Research Center of Massive Language Information Processing and Cloud Computing Applications Bijing Institute of Technology Beijing China Science and Technology on Special System Simulation Laboratory Beijing Simulation Center China

ISBN: (纸本)9781509035069

Modeling and simulation has received significant attention over the years for its ability to understand and transform real world systems. With the increasing of complexity and scale of the simulated systems, more computational power is required. Parallel discrete event simulation(PDES) is an important technology to integrate parallel computing into modeling and simulation. Time synchronization algorithm is the core of PDES, which determines the correctness and efficiency of the simulation. Traditional conservative synchronization algorithms, such as Chandy-Misra-Bryant(CMB) null message algorithm, use null message to synchronize the state and avoid deadlock among logical processes(LPs). But hybrid simulation systems contains both continuous models and discrete models, continuously changing states is typically modeled using time steps in hybrid simulation system. Therefore, hybrid systems with large disparity in time step are very common in real simulation systems. However, current conservative synchronization algorithm has not been fully investigated for this type of simulation systems and will generate massive synchronization messages. This paper proposed an enhanced null message algorithm for hybrid simulation systems with large disparity in time step. It extended the CMB algorithm by adding simulation time in null message request messages. This could significantly reduce the number of synchronization messages. Experimental results showed that our synchronization algorithm outperforms the CMB algorithm by reducing a great number of synchronization messages.

关键词： Synchronization Computational modeling Atmospheric modeling Adaptation models System recovery Ports (Computers) Complex systems

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：