检索结果-内蒙古大学图书馆

Annual meeting of the Association for Computational Linguistics

作者： Yang Feng Dongdong Zhang Mu Li Ming Zhou Qun Liu University of Sheffield Sheffield UK Microsoft Research Asia Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences

ISBN: (纸本)9781622761715

We present a hierarchical chunk-to-string translation model, which can be seen as a compromise between the hierarchical phrase-based model and the tree-to-string model, to combine the merits of the two models. With the help of shallow parsing, our model learns rules consisting of words and chunks and meanwhile introduce syntax cohesion. Under the weighed synchronous context-free grammar defined by these rules, our model searches for the best translation derivation and yields target translation simultaneously. Our experiments show that our model significantly outperforms the hierarchical phrase-based model and the tree-to-string model on English-Chinese Translation tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

${\rm S}^{3}{\rm MKL}$: Scalable Semi-Supervised Multiple Kernel Learning for Real-World Image Applications

引用

IEEE Transactions on Multimedia 2012年第4期14卷 1259-1274页

作者： Shuhui Wang Qingming Huang Shuqiang Jiang Qi Tian Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China Department of Computer Science University of Texas-San Antonio San Antonio TX USA

We study the visual learning models that could work efficiently with little ground-truth annotation and a mass of noisy unlabeled data for large scale Web image applications, following the subroutine of semi-supervised learning (SSL) that has been deeply investigated in various visual classification tasks. However, most previous SSL approaches are not able to incorporate multiple descriptions for enhancing the model capacity. Furthermore, sample selection on unlabeled data was not advocated in previous studies, which may lead to unpredictable risk brought by real-world noisy data corpse. We propose a learning strategy for solving these two problems. As a core contribution, we propose a scalable semi-supervised multiple kernel learning method (S 3 MKL) to deal with the first problem. The aim is to minimize an overall objective function composed of log-likelihood empirical loss, conditional expectation consensus (CEC) on the unlabeled data and group LASSO regularization on model coefficients. We further adapt CEC into a group-wise formulation so as to better deal with the intrinsic visual property of real-world images. We propose a fast block coordinate gradient descent method with several acceleration techniques for model solution. Compared with previous approaches, our model better makes use of large scale unlabeled images with multiple feature representation with lower time complexity. Moreover, to address the issue of reducing the risk of using unlabeled data, we design a multiple kernel hashing scheme to identify the “informative” and “compact” unlabeled training data subset. Comprehensive experiments are conducted and the results show that the proposed learning framework provides promising power for real-world image applications, such as image categorization and personalized Web image re-ranking with very little user interaction.

关键词： Kernel Visualization Training Data models Noise measurement Training data Learning systems

来源：评论

学校读者我要写书评

暂无评论

Memory performance prediction of web server applications based on grey system theory

引用

14th Asia Pacific Web Technology Conference, APWeb 2012

作者： Huang, Faliang Zhang, Shichao Yuan, Changan Zhong, Zhi Faculty of Software Fujian Normal University Fuzhou 350007 China Faculty of Engineering and Information Technology University of Technology Broadway P.O. Box 123 Sydney NSW 2007 Australia Science Computing and Intelligent Information Processing GuangXi Higher Education Key Laboratory Nanning 530023 China

ISBN: (纸本)9783642292521

With the success of internet, recently more and more companies start to run web-based business. While running e-business sites, many companies have encountered unexpected degeneration of their web server applications performance, which may lead to loss of customers. Many managers wish to have a decision-support tool that cancan answer such questions, such as "will my web server applications performance degenerate?", and "what are the main reasons of the degenerations?". In this paper we first propose a new memory performance prediction model of web server applications based on grey system theory. And then, a software system "Memory Performance Manager" (MPM) is developed for predicting memory performance of the web server applications. Massive experiments demonstrate that the effectiveness of MPM's in predicting web server memory performances. © 2012 Springer-Verlag Berlin Heidelberg.

关键词： Managers

来源：评论

学校读者我要写书评

暂无评论

A novel online event analysis framework for micro-blog based on incremental topic modeling

A novel online event analysis framework for micro-blog based...

引用

13th ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed computing, SNPD 2012

作者： Ma, Huifang Wang, Bo Li, Ning Department of Computer Science Northwest Normal University Lanzhou Gansu 730070 China Department of Military Information Management Shanghai Branch PLA Nanjing Institute of Politics Shanghai 200433 China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences 100190 Beijing China

ISBN: (纸本)9780769547619

In this paper, we present a scalable implementation of a topic modeling (Adaptive Link-IPLSA) based method for online event analysis, which summarize the gist of massive amount of changing tweets and enable users to explore the temporal trends in topics. This model also can simultaneously maintain the continuity of the latent semantics to better capture the time line development of events. With the help of this model, users can quickly grasp major topics in these twitters. The preliminary results show that our method leads to more balanced and comprehensive improvement for online event detection compared to benchmark approaches. Additionally our algorithm is computationally feasible in near real-time scenarios making it an attractive alternative for capturing the rapidly changing dynamics of microblogs. © 2012 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

EXIT Chart-Based Side Information Refinement for Wyner-Ziv Video Coding

EXIT Chart-Based Side Information Refinement for Wyner-Ziv V...

引用

Data Compression Conference (DCC)

作者： Wen Ji Pascal Frossard Yiqiang Chen Beijing Key Laboratory of Mobile Computing and Pervasive Device Institute of Computing Technology Chinese Academy of Sciences China Signal Processing Laboratory (LTS4) Ecole Polytechnique Fédéral de Lausanne Switzerland

This paper focuses on side information (SI) refinement in Wyner-Ziv video coding and proposes to exploit the intrinsic property of channel coding for improving the joint decoding performance. In this paper, we propose to use syndrome and information bits from the encoder to help the decoder in refining the SI. We use extrinsic information transfer (EXIT) chart analysis to deduce the mutual information variation in LDPC iterative decoding during the SI refinement process. The objective is to obtain the same decoding quality under lower coding rates. Simulation results demonstrate the effectiveness of the proposed solution.

关键词： Silicon Decoding Iterative decoding Video coding Correlation Mutual information

来源：评论

学校读者我要写书评

暂无评论

Towards collecting commonsense knowledge with hint questions

Towards collecting commonsense knowledge with hint questions

引用

2012 8th International Conference on Semantics, Knowledge and Grids, SKG 2012

作者： Zang, Liangjun Cao, Cong Cao, Cungen Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences No.6 Kexueyuan South Road Hai Dian Distinct Beijing 100190 China Graduate University Chinese Academy of Sciences No.19 Yu Quan Road Shi Jing Shan Distinct Beijing 100049 China

ISBN: (纸本)9780769547947

Accumulating commonsense knowledge (CK) has proven very useful for many natural language processing tasks. So far the most reliable way of acquisition is still relying on knowledge contributors to offer CK. Unfortunately, knowledge contributors often fail to think of much CK because this knowledge is usually taken for granted by people. In this paper, we explore the issue of reminding knowledge contributors by means of hint questions. Specifically, we investigate what kinds of CK are easily ignored, analyze its reason and suggest possible solution. We hope to sketch sources of the research problem and suggest possible ways forward. © 2012 IEEE.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Challenges and Opportunities for Trust Management in Crowdsourcing

Challenges and Opportunities for Trust Management in Crowdso...

引用

IEEE WIC ACM International Conference on Web Intelligence (WI)

作者： Han Yu Zhiqi Shen Chunyan Miao Bo An School of Computer Engineering Nanyang Technological University Singapore The Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China

Crowd sourcing (CS) systems offer a new way for businesses and individuals to leverage on the power of mass collaboration to accomplish complex tasks in a divide-and-conquer manner. In existing CS systems, no facility has been provided for analyzing the trustworthiness of workers and providing decision support for allocating tasks to workers, which leads to high dependency of the quality of work on the behavior of workers in CS systems as shown in this paper. To address this problem, trust management mechanisms are urgently needed. Traditional trust management techniques are focused on identifying the most trustworthy service providers (SPs) as accurately as possible. Little thoughts were given to the question of how to utilize these SPs due to two common assumptions: 1) an SP can serve an unlimited number of requests in one time unit, and 2) a service consumer (SC) only needs to select one SP for interaction to complete a task. However, in CS systems, these two assumptions are no longer valid. Thus, existing models cannot be directly used for trust management in CS systems. This paper takes the first step towards a systematic investigation of trust management in CS systems by extending existing trust management models for CS trust management and conducting extensive experiments to study and analyze the performance of various trust management models in crowd sourcing. In this paper, the following key contributions are made. We 1) propose extensions to existing trust management approaches to enable them to operate in CS systems, 2) design a simulation test-bed based on the system characteristics of Amazon's Mechanical Turk (AMT) to make evaluation close to practical CS systems, 3) discuss the effect of incorporating trust management into CS system on the overall social welfare, and 4) identify the challenges and opportunities for future trust management research in CS systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Generalized two-dimensional FLD with assembled matrix distance metric for face recognition

引用

Journal of Information and Computational Science 2012年第12期9卷 3439-3446页

作者： Shi, Yuexiang Gong, Ping Tao, Zhiping Key Laboratory of Intelligent Computing and Information Processing of Ministry of Education School of Information Engineering Xiangtan University Xiangtan 411105 China School of Information Engineering Xiangtan University Xiangtan 411105 China

Two-dimensional fisher linear discriminant analysis (2DFLD or 2DLDA) has attracted much attention from researchers recently for the advantages over the singularity problem and the computational cost. Recent research on 2DLDA-based methods has mainly concentrated on two issues, namely: 1) feature extraction and 2) classification. In this paper, we further analyze these two problems simultaneously by using G-2DFLD supplemented with an Assembled Matrix Distance (AMD) metric. For feature extraction, G-2DFLD is proposed, which can be used for image feature extraction by reducing the dimensionality in both horizontal and vertical directions. For classification, an AMD metric is presented to calculate the distance between two feature matrices, then, the nearest neighbor classifier is used for face recognition. To test this new method, ORL face database is used and the results show the efficiency of G-2DFLD with AMD metric in face recognition. © 2012 Binary Information Press.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Properties and axiomatization of fuzzy rough sets based on fuzzy covering

Properties and axiomatization of fuzzy rough sets based on f...

引用

International Conference on Machine Learning and Cybernetics (ICMLC)

作者： Ming-Fen Wu Hao-Han Han Yan-Fei Si School of Computer Science Wuyi University Guangdong China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China

Some expanded fuzzy rough sets models have been investigated to handle fuzzy databases with uncertain, imprecise and incomplete real-valued information. In this paper, we make further research on fuzzy rough sets models in fuzzy environment, and we generalize rough fuzzy sets model based on a covering to fuzzy rough sets model based on a fuzzy covering. The lower and upper approximations of fuzzy subsets are defined based on a fuzzy covering, and basic properties are investigated. Then, the axiom definition of the lower approximation operator is given. It is shown that the rough fuzzy sets model based on a covering is a special instance of the fuzzy rough sets model based on a fuzzy covering.

关键词： Abstracts

来源：评论

学校读者我要写书评

暂无评论

Formal concept analysis in hybrid relational databases

引用

International Review on Computers and Software 2012年第6期7卷 2904-2910页

作者： Lei, Yuxia Sui, Yuefei Cao, Baoxiang School of Computer Science Qufu Normal University Rizhao 276826 Shandong China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China

Relational Database Model (RDM) has been proven to be a very useful data-storage technique. As information is stored as data in relational databases, the induction of concepts from data is a pivotal topic in the data mining field. Formal Concept Analysis (FCA) turns out to be a perfect instrument for a meaningful and conceptual exploration of the stored data. However, many databases in practice are often hybrid ones. This paper proposes an approach towards extracting concepts from hybrid relations, which preserve the RDM and allow complex values such as relations in columns for relations. The method can be described as follows: the sub-tuples/subattributes pairs are regard as new attributes and inserted into a main relation, and we further obtain a new complex relation called a target relation from which concepts can be extracted. The insertion operation is semantics-preserving and extent-preserving. In addition, this paper analyzed the structural connections among concept lattices in detail, which can be used to construct a complex concept lattice from simple ones. The method may reduce time complexity of constructing the concept lattice of a complex relation. © 2012 Praise Worthy Prize S.r.l. - All rights reserved.

关键词： Formal concept analysis

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：