检索结果-内蒙古大学图书馆

Large-scale influence maximization with the influence maximization benchmarker suite 4

Large-scale influence maximization with the influence maximi...

4th ACM SIGMOD Joint International Workshop on Graph Data Management Experiences and systems and Network Data Analytics, GRADES-NDA 2021

作者： Geppert, Heiko Bhowmik, Sukanya Rothermel, Kurt Institute for Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781450384773

Maximizing the influence of a fixed size seed set in a graph was investigated intensively in the past decade. Two very relevant questions are 1) how to solve the influence maximization problem on very large graphs within short time and 2) how to compare possible findings with the current state-of-the-art in a fair manner. To solve the first problem, proxy-based influence maximization strategies emerged. However, today's graphs became too large to be solved quickly for many well-established proxy strategies, since they do not scale to such large graphs. In this paper we propose 1) a novel update scheme for iterative influence maximization strategies named Update Approximation (UA) capable of large influence spreads within a few seconds on billion-scale graphs. Further, we present 2) a generic benchmark suite (Influence Maximization Benchmarker - - IMB) to implement and evaluate influence maximization strategies, alongside with implementations for several established strategies. IMB allows for easy to use benchmarks for further research by the community. © 2021 ACM.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

AutoTuneTMP: Auto-tuning in C++ with runtime template metaprogramming 32

AutoTuneTMP: Auto-tuning in C++ with runtime template metapr...

引用

32nd IEEE International parallel and distributed Processing Symposium Workshops, IPDPSW 2018

作者： Pfander, David Brunn, Malte Pfluger, Dirk Institute for Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781538655559

Careful tuning of code is crucial to obtain near-optimal runtime performance on the hardware at hand. How-ever, current hardware platforms pose plenty of challenges: multiple levels of cache, many cores and increasingly wider vector units render manual tuning time-consuming and cumbersome. Thus, algorithms have to be able to automatically tune themselves to the underlying hardware platform to maximize performance and to ensure performance portability. In this work, we present AutoTuneTMP, a novel C++-based auto-tuning library. Its unique strength is to combine template metaprogramming and just-in-time (JIT) compilation so that templates can be instantiated at runtime. We use this runtime metaprogramming approach to provide an extensible set of parameterized template-based optimizations and data structures for writing tunable kernels. Together with convenience functionality for parameter tuning, this lightweight approach can be used to implement algorithms with near-optimal and portable performance. We demonstrate the applicability, usefulness and performance of AutoTuneTMP at the example of matrix multiplication. It is well-suited as a demonstrator, being a well-known compute kernel and exhibiting a rather large set of (nine) tunable parameters to be optimized. With simple search algorithms, AutoTuneTMP is able to automatically achieve up to 90% of the peak performance on a Xeon Silver 4116 processor and up to 34% on a Xeon Phi 7210. Starting from an already parallelized and vectorized baseline, we obtained a speedup of up to 3.1x on the Xeon Silver platform and up to 6.7x on the Xeon Phi platform. © 2018 IEEE.

关键词： Silver

来源：评论

学校读者我要写书评

暂无评论

PreMatch: A Prediction-based Scheduler for Dataflow Asynchronous Processing on the Computing Continuum 25

PreMatch: A Prediction-based Scheduler for Dataflow Asynchro...

引用

Proceedings of the 40th ACM/SIGAPP Symposium on Applied Computing

作者： Narges Mehran Zahra Najafabadi Samani Samira Afzal Radu Prodan Frank Pallas Peter Dorfinger Intelligent Connectivity Group Salzburg Research Forschungsgesellschaft mbH Salzburg Salzburg Austria Faculty of Digital and Analytical Sciences Paris Lodron University of Salzburg Salzburg Salzburg Austria Distributed and Parallel Systems Group University of Innsbruck Innsbruck Tyrol Austria Faculty of Electrical Engineering Mathematics and Computer Science University of Twente Enschede Overijssel Netherlands Institute of Information Technology Alpen-Adria-University of Klagenfurt Klagenfurt Carinthia Austria

ISBN: (纸本)9798400706295

The popularity of asynchronous data exchange patterns has recently increased, as evidenced by 23% of the communication between microservices in an Alibaba trace analysis. Such workloads necessitate methods for reducing dataflow processing and completion time by forecasting the future requirements of their microservices and (re-)scheduling them. Therefore, we investigate a prediction-based scheduling method of asynchronous dataflow processing applications by considering the stochastic changes due to dynamic user requirements. We present a microservice scaling and scheduling method named PreMatch combining a machine learning (ML) prediction strategy based on gradient boosting with ranking and game theory matching scheduling principles. Firstly, PreMatch predicts the number of microservice replicas, and then, the ranking method orders the microservice replicas and devices based on microservice and transmission times. Thereafter, the PreMatch schedules microservice replicas requiring dataflow processing on computing devices. Experimental analysis of the PreMatch method shows lower completion times on average 13% compared to a related prediction-based scheduling method.

关键词： microservice

来源：评论

学校读者我要写书评

暂无评论

The power of software-defined networking: Line-rate content-based routing using OpenFlow

The power of software-defined networking: Line-rate content-...

引用

7th Workshop on Middleware for Next Generation Internet Computing, MW4NG 2012 - Co-located with the 13th ACM/IFIP/USENIX International Middleware Conference, Middleware 2012

作者： Koldehofe, Boris Dürr, Frank Tariq, Muhammad Adnan Rothermel, Kurt Institute of Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781450316071

A lot of research effort has been invested to support efficient content-based routing. Nevertheless, practitioners often fall back to far less expressive communication paradigms like multicast groups. The benefits of content-based routing in minimizing bandwidth consumption are often rendered useless by simpler communication paradigms that rely on line-rate processing of data packets at the switches of the network providers. Contrary content-based routing protocols face the inherent overhead in matching the content of events against subscriptions leading to far lower throughput rates and higher end-to-end delays. However, recent trends in networking such as software defined networking in combination with network virtualization have tremendous potential to change the picture. In our opinion this will significantly increase acceptance of sophisticated middleware like content-based routing in the future. To support our claims we outline in this paper a reference architecture that may be used to build middleware for Future Internet applications. Furthermore, we provide a solution for realizing content-based routing at line-rate relying on this reference architecture and illustrate research problems that need to be addressed. © 2012 ACM.

关键词： Software defined networking

来源：评论

学校读者我要写书评

暂无评论

A new approach for establishing pairwise keys for securing wireless sensor networks 05

A new approach for establishing pairwise keys for securing w...

引用

3rd ACM International Conference on Embedded Networked Sensor systems, SenSys 2005

作者： Wacker, Arno Knoll, Mirko Heiber, Timo Rothermel, Kurt Dept. of Distributed Systems Institute of Parallel and Distributed Systems Universiẗat Stuttgart Stuttgart Germany

ISBN: (纸本)159593054X

Wireless sensor networks based on highly resource-constrained devices require symmetric cryptography in order to make them secure. Integral to this is the exchange of unique symmetric keys between two devices. In this paper, we propose a novel decentralized key exchange protocol that guarantees the confidentiality of a key exchange even if an attacker has compromised some of the devices in the network. A central objective of the protocol design was to minimize resource consumption on the individual devices. We evaluate the resource requirements of our protocol in terms of memory requirements, CPU usage and network traffic both through theoretical analysis and through simulations. Categories and Subject Descriptors D.4.6 [Operating systems]: Security and Protection, Cryptographic controls General Terms Algorithms, Security. © 2005 ACM.

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

Benefits of integrating meta data into a context model

Benefits of integrating meta data into a context model

引用

Third IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom 2005 Workshops

作者： Hönle, Nicola Käppeler, Uwe-Philipp Nicklas, Daniela Schwarz, Thomas Grossmann, Matthias Institute of Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)0769523005

Meta data - data about data - improves the value of the operational data by giving applications and users additional information on the data's origin, its precision or stateness. We outline the benefits of modeling meta data in context models: it can be used for resource finding, enhanced data selection, trust and data quality issues and sensor fusion. We show how meta data included into an object-based context model influences the data modeling and the selection process in the query language. Finally, we describe our implementation of the presented functionality in the Nexus platform. © 2005 IEEE.

关键词： Metadata

来源：评论

学校读者我要写书评

暂无评论

Flexible application-aware approximation for modern distributed graph processing frameworks 5

Flexible application-aware approximation for modern distribu...

引用

5th ACM SIGMOD Joint International Workshop on Graph Data Management Experiences and systems and Network Data Analytics, GRADES-NDA 2022, co-located with ACM SIGMOD Conference

作者： Schramm, Michael Bhowmik, Sukanya Rothermel, Kurt Institute for Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781450393843

The interest in the ability of processing data that has an underlying graph structure has grown in the recent past. This has led to the development of many distributed graph processing systems. However, due to rapidly growing amount of data, e.g., web graphs and social graphs, even such distributed graph processing frameworks end up requiring several minutes or even several hours to execute popular graph algorithms. This leads to the question: do we always need to know the exact answer for a large graph? The aforementioned modern distributed graph processing frameworks execute graph algorithms by exchanging messages between vertices. This paper introduces a novel message-dropping approach for approximation in these frameworks. As dropping messages would result in degradation of quality of result, our objective is to drop messages that have least adverse impact on quality. More specifically, we propose an application-aware approach that dynamically drops messages at runtime. We evaluate the effects of our approach for the PageRank algorithm on several representative real-world web graphs and compare its performance to that of existing approximation techniques for modern frameworks.. © 2022 ACM.

关键词： Drops

来源：评论

学校读者我要写书评

暂无评论

SplitPass： A Mutually Distrusting Two-Party Password Manager

引用

Journal of Computer Science & Technology 2018年第1期33卷 98-115页

作者： Yu-Tao Liu Dong Du Yu-Bin Xia Hai-Bo Chen Bin-Yu Zang Zhenkai Liang Institute of Parallel and Distributed Systems Shanghai Jiao Tong University Shanghai 200240 China School of Computing National University of Singapore Singapore 117417 Singapore

Using a password manager is known to be more convenient and secure than not using one, on the assmnption that the password manager itself is safe. However recent studies show that most popular password managers have security vulnerabilities that may be fooled to leak passwords without users＇ awareness. In this paper, we propose a new password manager, SplitPass, which vertically separates both the storage and access of passwords into two mutually distrusting parties. During login, all the parties will collaborate to send their password shares to the web server, but none of these parties will ever have the complete password, which significantly raises the bar of a successful attack to compromise all of tile parties. To retain transparency to existing applications and web servers, SplitPass seamlessly splits the secure sockets layer （SSL） and transport layer security （TCP） sessions to process on all parties, and makes the joining of two password shares transparent to the web servers. We have implemented SplitPass using an Android phone and a cloud assistant and evaluated it using 100 apps from top free apps in the Android official market. The evaluation shows that SplitPass securely protects users＇ passwords, while incurring little performance overhead and power consumption.

关键词： password manager privacy protection mobile-cloud system

来源：评论

学校读者我要写书评

暂无评论

SDOS: Using trusted platform modules for secure cryptographic deletion in the swift object store 20

SDOS: Using trusted platform modules for secure cryptographi...

引用

20th International Conference on Extending Database Technology, EDBT 2017

作者： Waizenegger, Tim Wagner, Frank Mega, Cataldo University of Stuttgart Institute for Parallel and Distributed Systems Germany

ISBN: (纸本)9783893180738

The secure deletion of data is becoming increasingly important to individuals, corporations as well as governments. Recent advances in worldwide laws and regulations now require secure deletion for sensitive data in certain industries. Data leaks in the public and private sector are commonplace today, and they often reveal data which was supposed to be deleted. Secure deletion describes any mechanism that renders stored data unrecoverable, even through forensic means. In the past this was achieved by destroying storage media or overwriting storage sectors. Both of these mechanisms are not well suited to today’s multi-tenant cloud storage solutions. Cryptographic deletion is a suitable candidate for these services, but a research gap still exists in applying cryptographic deletion to large cloud storage services. Cloud providers today rarely offer storage solutions with secure deletion for these reasons. In this Demo, we present a working prototype for a cloud storage service that offers cryptographic deletion with the following two main contributions: A key-management mechanism that enables cryptographic deletion an on large volume of data, and integration with Trusted Platform Modules (TPM) for securing master keys. © 2017, Copyright is with the authors.

关键词： Information management

来源：评论

学校读者我要写书评

暂无评论

The tensor flow partitioning and scheduling problem: It's the critical path! 1

The tensor flow partitioning and scheduling problem: It's th...

引用

1st Workshop on distributed Infrastructures for Deep Learning, DIDL 2017

作者： Mayer, Ruben Mayer, Christian Laich, Larissa Institute of Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781450351690

State-of-The-Art data flow systems such as TensorFlow impose iterative calculations on large graphs that need to be partitioned on heterogeneous devices such as CPUs, GPUs, and TPUs. However, partitioning can not be viewed in isolation. Each device has to select the next graph vertex to be executed, i.e., perform local scheduling decisions. Both problems, partitioning and scheduling, are NP-complete by themselves but have to be solved in combination in order to minimize overall execution time of an iteration. In this paper, we propose several heuristic strategies to solve the partitioning and scheduling problem in TensorFlow. We simulate the performance of the proposed strategies in heterogeneous environments with communication-intensive workloads that are common to TensorFlow. Our findings indicate that the best partitioning and scheduling heuristics are those that focus on minimizing the execution time of the critical path in the graph. Those strategies provide a speed-up of up to 4 times in comparison to strategies that are agnostic to the critical path, such as hash-based partitioning and FIFO scheduling. © 2017 Copyright held by the owner/author(s).

关键词： Tensors

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：