检索结果-内蒙古大学图书馆

2017 IEEE 2nd International Conference on Big Data Analysis（ICBDA 2017）

作者： Shangru Zhong Kai Lei Xiaohui Huang Jincheng Wu Institute of Big Data Technologies Shenzhen Key Lab for Cloud Computing Technology & Applications School of Electronic and Computer Engineering Peking University

ISBN: (纸本)9781509036202

With tags widely used in organizing and searching contents in massive data era,how to automatically generate appropriate tags of resource for users became a hot issue on social networks *** recommendation for text resource can be modeled as a keyword extraction problem,hence topic modeling such as LDA which extracts latent semantic topics from text is suitable for tag ***,latent topics are too coarse-grained to describe ***,LDA trains corpus globally without considering context ***,topics generated are difficult to be quantifiably *** problems lead to the poor quality of tag recommendation in topic model based *** this paper,we propose topic representation method,which introduces embedding semantic representation into topic *** results of evaluation on real social networks show that the proposed method improves the quality of tag recommendation for Chinese text resource,when comparing with traditional LDA-based method,which demonstrates the effectiveness of modifying topic modeling.

关键词： Tag recommendation topic model LDA word embedding

来源：评论

学校读者我要写书评

暂无评论

Syndromes Diagnostic Model for Coronary Artery Disease(CAD):An Improved Na?ve Bayesian Classification Model Based on Attribute Relevancy 2

Syndromes Diagnostic Model for Coronary Artery Disease(CAD):...

引用

2017 IEEE 2nd International Conference on Big Data Analysis（ICBDA 2017）

作者： Kai Lei Lizhu Zhang Ying Shen Xiaohui Huang Jincheng Wu Institute of Big Data Technologies Shenzhen Key Lab for Cloud Computing Technology and Applications School of Electronic and Computer Engineering Peking University

ISBN: (纸本)9781509036202

Clinical decision support is a probabilistic and quantitative method learning used for problem modeling in situations of ***,mathematical methods such as probability learning and statistical analysis is increasingly used in the medical field to assure the accuracy of the diagnoses and optimize the *** study describes the design and implementation of an improved weighted Na?ve Bayes classifier based on the idea that there is a certain level of dependency between *** the coronary artery disease,clinical data and patients' medical records were used to train the improved Na?ve Bayes classifier to compute the possible disease manifestations,the priori probability and the posterior probability and to carry out the disease ***,we were able to improve the precision of the improved weighted Na?ve Bayes classifier to 79%(increase of 12% compared to Na?ve Bayes).The suitability of the improved Na?ve Bayes classifiers for carrying out diagnosis allows them to be utilized to distinguish syndromes and signs based on unobserved evidence.

关键词： Na?ve Bayes decision support diagnostic classification attribute relevancy negative symptoms coronary artery disease

来源：评论

学校读者我要写书评

暂无评论

Large-scale wire-speed multicast switching structure based on multipath self-routing switching structure and implemented on FPGA

Large-scale wire-speed multicast switching structure based o...

引用

3rd International Conference on Materials Science and Information technology, MSIT 2013

作者： Cui, Kai Li, Hui Zhu, Zhi Pu Chen, Fu Xing Shenzhen Eng. Lab of Converged Networks Technology Shenzhen Key Lab of Cloud Computing Tech. and App Shenzhen Graduate School Peking University Shenzhen China

ISBN: (纸本)9783037857700

Ensuring high quality of service (QoS) of multicast video stream over next generation network is a challenging issue, and how to implement the wire-speed multicast with hardware logical support in the network nodes of every hierarchy is a key solution to achieve high QoS multicast. Currently, the multicast packets are processed in this way, in which they are copied and then scheduled by ports, lastly, sent respectively. But this approach cannot ensure high QoS in realtime applications. Moreover, the traditional hardware solutions do not possess excellent large-scale scalability owning to their own bottlenecks. In this project, we have constructed a wire-speed multicast switching structure based on Multipath Self-routing Switching Structure, and implemented it on a Stratix IV FPGA chip. Additionally, we have designed the signaling system and control mechanism to support the process of self-routing and wire-speed fan-out copy of multicast packets. © (2013) Trans Tech Publications, Switzerland.

关键词： Wire

来源：评论

学校读者我要写书评

暂无评论

BCStore: Bandwidth-Efficient in-memory kv-store with batch coding 33

BCStore: Bandwidth-Efficient in-memory kv-store with batch c...

引用

33rd International Conference on Massive Storage Systems and technology, MSST 2017

作者： Li, Shenglong Zhang, Quanlu Yang, Zhi Dai, Yafei Peking University China Institute of Big Data Technologies Shenzhen Key Lab for Cloud Computing Technology and Applications China

In-memory key-value store KV-store is a crucial building block for many systems including large-scale web services and databases. Recent research works pay close attention to data availability of in-memory KV-store. Replication is a common solution to achieve data availability, but the significant memory consumption makes it not applicable for memoryintensive applications. Integrating erasure coding into inmemory KV-store is an emerging approach to achieve memory efficiency. However, traditional in-place update mechanism of erasure coding incurs significant bandwidth cost for writeintensive workload due to frequent broadcast of data update. In this paper, we build BCStore, a bandwidth-efficient inmemory KV-store with erasure coding. We propose a novel batch coding mechanism to optimize the bandwidth cost, and verify its feasibility and efficiency through theoretical analysis. To recycle the memory space due to delete and update operations, we design an efficient move-based garbage collection (GC) mechanism with a novel data layout called virtual coding stripe. Besides, BCStore guarantees consistent data read and write under these new designs, and supports fast online data recovery. We evaluate the performance of BCStore under various workloads. Experimental results show that BCStore can save up to 41% memory compared with replication, and achieve up to 2.4x throughput improvement and 51% bandwidth saving compared to erasure coding with in-place update. © 33rd International Conference on Massive Storage Systems and technology, MSST 2017. All rights reserved.

关键词： Bandwidth

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Computational Feature Selection Method for Medical Synonym Identification via Bilingualism and Multi-corpus Training 2

An Enhanced Computational Feature Selection Method for Medic...

引用

2017 IEEE 2nd International Conference on Big Data Analysis（ICBDA 2017）

作者： Kai Lei Shangchun Si Desi Wen Ying Shen Institute of Big Data Technologies Shenzhen Key Lab for Cloud Computing Technology and Applications School of Electronic and Computer Engineering(SECE) Peking University

ISBN: (纸本)9781509036202

Medical synonym identification has been an important part of medical natural language processing(NLP).However,in the field of Chinese medical synonym identification,there are problems like low precision and low recall *** solve the problem,in this paper,we propose a method for identifying Chinese medical *** first selected 13 features including Chinese and English *** we studied the synonym identification results of each feature alone and different combinations of the *** the comparison among identification results,we present an optimal combination of features for Chinese medical synonym *** show that our selected features have achieved 97.37% precision rate,96.00% recall rate and 97.33% F1 score.

关键词： medical synonym feature selection bilingual multi-corpus

来源：评论

学校读者我要写书评

暂无评论

DRWS: A model for learning distributed representations for words and sentences

Lecture Notes in Computer Science (including subseries Lectu...

引用

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2014年 8862卷 196-207页

作者： Yan, Chunwei Zhang, Fan Huang, Lian’en The Shenzhen Key Lab for Cloud Computing Technology and Application Peking University Shenzhen Graduate School ShenzhenGuangdong China

Vector-space distributed representations of words can capture syntactic and semantic regularities in language and help learning algorithms to achieve better performance in natural language processing tasks by grouping similar words. With progress of machine learning techniques in recent years, much attention has been paid on this field. However, many NLP tasks such as text summary and sentence matching treat sentences as atomic units. In this paper, we introduce a new model called DRWS which can learn distributed representations for words and variable-length sentences. Feature vectors for words and sentences are learned based on their probability of co-occurrence between words and sentences using a neural network. To evaluate feature vectors learned by our model, we applied our model on the tasks of detecting word similarity and text summarization. Extensive experiments demonstrate the effectiveness of our proposed model in learning vector representations for words and sentences. © Springer International Publishing Switzerland 2014.

关键词： Vectors

来源：评论

学校读者我要写书评

暂无评论

On the routing of wide-sense circuit based on algebraic switching fabric 9

On the routing of wide-sense circuit based on algebraic swit...

引用

9th International ICST Conference on Testbeds and Research Infrastructures for the Development of Networks and Communities, TridentCom 2014

作者： Zhan, Qian Li, Hui Ma, L. Lv, Shijie Shenzhen Eng. Lab of Converged Networks Technology Shenzhen Key Lab of Cloud Computing Tech. and App Shenzhen Graduate School Peking University Shen Zhen China

ISBN: (纸本)9783319133256

In order to ensure high quality of service for Next Generation Network, we focus our study on the Wide-Sense Circuit proposed in Flexible Architecture of Reconfigurable Infrastructure. First we construct the functional structure of Wide-Sense Circuit and then explore the switching mechanism and module for it, which is called the Multipath Self-routing Switching Mechanism. Its detailed working process includes traffic classification, establishment and adjustment of Wide-Sense Circuit and data forwarding three parts. For underlying data forwarding, we introduce an innovative Load-Balanced Multipath Self-routing Switching Architecture and start on the implementation on an Altera StratixIV FPGA. The inspiring test results prove that our theory and practiceguarantee the high communication transmission quality for Wide-Sense Circuit. © Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2014.

关键词： Timing circuits

来源：评论

学校读者我要写书评

暂无评论

WordTopic-MultiRank: A New Method for Automatic keyphrase Extraction 6

WordTopic-MultiRank: A New Method for Automatic Keyphrase Ex...

引用

6th International Joint Conference on Natural Language Processing, IJCNLP 2013

作者： Zhang, Fan Huang, Lian'en Peng, Bo Shenzhen Key Lab for Cloud Computing Technology and Applications Peking University Shenzhen Graduate School Shenzhen518055 China Institute of Network Computing and Information Systems Peking University Beijing100871 China

ISBN: (纸本)9784990734800

Automatic keyphrase extraction aims to pick out a set of terms as a representation of a document without manual assignment efforts. Supervised and unsupervised graph-based ranking methods have been studied for this task. However, previous methods usually computed importance scores of words under the assumption of single relation between words. In this work, we propose WordTopic-MultiRank as a new method for keyphrase extraction, based on the idea that words relate with each other via multiple relations. First we treat various latent topics in documents as heterogeneous relations between words and construct a multi-relational word network. Then, a novel ranking algorithm, named Biased-MultiRank, is applied to score the importance of words and topics simultaneously, as words and topics are considered to have mutual influence on each other. Experimental results on two different data sets show the outstanding performance and robustness of our proposed approach in automatic keyphrase extraction task. © IJCNLP *** right reserved.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

A general co/decoder of network coding in HDL

A general co/decoder of network coding in HDL

引用

IEEE International Symposium on Network Coding

作者： Zhang, Minglong Li, Hui Chen, Fuxing Hou, Hanxu An, Huiyao Wang, Wei Huang, Jiaqing Shenzhen Key Lab. of Cloud Computing Tech. and Application Shenzhen Graduate School Peking Univ. 518055 China Dept. of Electronic and Information Engineering Huazhong University of Science and Technology Wuhan 430074 China

ISBN: (纸本)9781612841403

This paper presents a practical and general coder and decoder of network coding (NC) with HDL (Hardware Description Language) logic for wire-speed nodes in multisource multicast networks. The NC coders apply random linear network coding (RLNC) and the decoders recover the original packets by Cramer's rule. All these mathematical operations are carried out in the Galois Field (2 8). The structures and algorithms of NC coder and decoder were designed in detail and implemented in HDL with NetFPGA boards provided by Stanford University. Comparing with traditional stored-and-forward mechanism, network emulations showed that networks with wire-speed NC coder and decoder nodes could achieve the capacity bound of max-flow min-cut theorem in case of bottlenecks, and the end-to-end delay was guaranteed on a small constant. © 2011 IEEE.

关键词： Network coding

来源：评论

学校读者我要写书评

暂无评论

Open domain question answering with character-level deep learning models

Open domain question answering with character-level deep lea...

引用

作者： Lei, Kai Deng, Yang Zhang, Bing Shen, Ying Institute of Big Data Technologies Shenzhen Key Lab for Cloud Computing Technology and Applications School of Electronic and Computer Engineering Peking University Shenzhen518055 China

ISBN: (纸本)9781538636749

Single-relation factoid question answering (QA) is strongly supported by rich sources of facts from knowledge bases (KB). However, there are many irrelevant information in questions and overwhelming number of facts in knowledge bases, making it difficult to capture goal entity and relation involved in a question. In order to settle these issues, firstly, a state-of-the-art sequence tagging model (BiLSTM-CRF) is adopted to detect the entity mention in a question. Then, we propose a n-gram match (NGM) algorithm with Chinese-specific rules and an attention-based siamese bidirectional long-short term memory (ASBLSTM) model to measure the lexical and semantic similarity between questions and candidate facts. Our whole method requires no hand-crafted template or feature engineering. In addition, character-level models are proved to be effective in solving the out of vocabulary (OOV) issue and improving the accuracy in Chinese KBQA task. Experiment results show that our system outperforms the best system with deep learning models in the KBQA share task of the Conference on Natural Language Processing and Chinese computing (NLPCC) 2016 and our system achieves an AverageF1 measure of 80.97% and 37.18% on test dataset in NLPCC 2016 and 2017 respectively. © 2017 IEEE.

关键词： Entity extraction Entity linking Knowledge base question answering Relation recognition

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：