检索结果-内蒙古大学图书馆

Ivannikov ISPRAS Open Conference (ISPRAS)

作者： Kirill Lukyanov Mikhail Drobyshevskiy Danil Shaikhelislamov Denis Turdakov Ivannikov Institute for System Programming of the Russian Academy of Sciences Moscow Russia Moscow Institute of Physics and Technology (National Research University) Moscow Russia

Social networks crawling is in the focus of active research the last years. One of the challenging task is to collect target nodes in an initially unknown graph given a budget of crawling steps. Predicting a node property based on its partially known neighbourhood is at the heart of a successful crawler. In this paper we adopt graph neural networks for this purpose and show they are competitive to traditional classifiers and are better for individual cases. Additionally we suggest a training sample boosting technique, which helps to diversify the training set at early stages of crawling and thus improves the predictor quality. The experimental study on three types of target set topology indicates GNN based approach has a potential in crawling task, especially in the case of distributed target nodes.

关键词： Training Heart Social networking (online) Network topology Crawlers Boosting Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Accelerated Methods with Compressed Communications for Distributed Optimization Problems under Data Similarity

arXiv

引用

arXiv 2024年

作者： Bylinkin, Dmitry Beznosikov, Aleksandr Moscow Institute of Physics and Technology Russia Ivannikov Institute for System Programming of the Russian Academy of Sciences Russia Sber AI Lab Skoltech Russia

In recent years, as data and problem sizes have increased, distributed learning has become an essential tool for training highperformance models. However, the communication bottleneck, especially for high-dimensional data, is a challenge. Several techniques have been developed to overcome this problem. These include communication compression and implementation of local steps, which work particularly well when there is similarity of local data samples. In this paper, we study the synergy of these approaches for efficient distributed optimization. We propose the first theoretically grounded accelerated algorithms utilizing unbiased and biased compression under data similarity, leveraging variance reduction and error feedback frameworks. In terms of communication time our theory gives O ( 1 + h M-1/4 + ω-1/2 ip δ/μ ) complexity for unbiased compressors and O ( 1 + β1/4 √ δ/μ ) for biased ones, where M is the number of computational nodes, β is the compression power, δ is the similarity measure and μ is the parameter of strong convexity of the objective. Our theoretical results are of record and confirmed by experiments on different average losses and datasets. Copyright © 2024, The Authors. All rights reserved.

关键词： Data compression

来源：评论

学校读者我要写书评

暂无评论

A Formal Model of XML Schema

A Formal Model of XML Schema

引用

International Conference on Data Engineering Workshops

作者： L. Novak A. Zamulin Institute for System Programming Russian Academy of Science Moscow Russia A.P. Ershov Institute of Informatics Systems Russian Academy of Science Novosibirsk Russia

The semantics of the core features of XML Schema in terms of XQuery 1.0 and XPath 2.0 data model algebraically defined is given. The database state is represented as a many sorted algebra whose sorts are sets of data type values and different kinds of nodes and whose operations are data type operations and node accessors. The values of some node accessors, such as "parent", "children" and "attributes", define a document tree with a definite order of nodes. The values of other node accessors help to make difference between kinds of nodes, learn the names, types and values associated with the corresponding document entities, etc., i.e., provide primitive facilities for a query language. As a result, a document can be easily mapped to its implementation in terms of nodes and accessors defined on them.

关键词： XML Data models Algebra Database languages Informatics Spatial databases Data engineering

来源：评论

学校读者我要写书评

暂无评论

On the possibility of provably secure obfuscating programs

引用

5th International Andrei Ershov Memorial Conference on Perspectives of system Informatics, PSI 2003

作者： Varnovsky, Nikolay P. Zakharov, Vladimir A. Laboratory for Cryptography Lomonosov State University MoscowRU-119899 Russia Lomonosov State University MoscowRU-119899 Russia Institute for System Programming Russian Academy of Sciences B. Kommunisticheskaya 25 Moscow109004 Russia

ISBN: (纸本)3540208135

By obfuscation we mean any efficient semantic-preserving transformation of computer programs aimed at bringing a program into such a form, which impedes the understanding of its algorithm and data structures or prevents the extracting of some valuable information from the plaintext of a program. The main difficulty in designing an effective program obfuscator is to guarantee security, i.e. to prove that no algorithm can break software protection in reasonable time. All obfuscation techniques and tools developed so far rely on the informal concept of security and therefore can’t be regarded as provably secure. In this paper we (1) introduce for the first time a formal information-theoretic definition of obfuscation security, (2) present a new obfuscation technique which takes advantage of cryptographic primitives (one-way functions, hard-core predicates), and (3) demonstrate, taking a conventional password identification scheme as a case study, how to prove security of the obfuscating transformations. © Springer-Verlag Berlin Heidelberg 2003.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Improving the Accuracy of Library Function Modeling in the Static Analyzer

Improving the Accuracy of Library Function Modeling in the S...

引用

Ivannikov ISPRAS Open Conference (ISPRAS)

作者： M. G. Biktimirov V. N. Ignatyev M. V. Belyaev Ivannikov Institute for System Programming of the Russian Academy of Sciences (ISP RAS) Lomonosov Moscow State University Moscow Russian Federation

Program's source code usually contains many calls of library functions without available sources, therefore there is no way to analyze them like user code. To achieve accuracy required by industrial static analyzer, various library modeling methods are used. The paper presents one of the methods used by the SharpChecker static source code analyzer for modeling library functions using C# specifications in the symbolic execution engine. The approach allows to store and explore predicates of events, important side-effects and other information useful for error detectors with reasonable accuracy. It makes it possible to model side-effects of user-defined lambda-functions passed to external methods as parameters, which is especially important for frequently used C# LINQ methods. The effectiveness of these improvements was evaluated on the set of open source projects containing 6 million lines of code.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RKDG method for 2D gas dynamics simulation on uniform rectangular meshes 10

RKDG method for 2D gas dynamics simulation on uniform rectan...

引用

10th All-Russian Conference on Irreversible Processes in Nature and Technology

作者： Korchagova, Victoria Fufaev, Ivan Lukin, Vladimir Marchevsky, Ilia Sautkina, Sofya Bauman Moscow State Technical University 2-nd Baumanskaya st. 5 Moscow105005 Russia Ivannikov Institute for System Programming of the RAS Alexander Solzhenitsyn st. 25 Moscow109004 Russia Keldysh Institute of Applied Mathematics RAS Miusskaya sq. 4 Moscow125047 Russia

This paper is devoted to a numerical simulation of 2D gas dynamics flows on uniform rectangular meshes using the Runge-Kutta-Discontinuous-Galerkin (RKDG) method. The RKDG algorithm was implemented with in-house C++ code based on the experience in the investigation of 1D case. The advantage of the RKDG method over the most popular finite volume method (FVM) is discussed: Three basis functions being applied in the framework of the RKDG approach lead to a considerable decrease of the numerical dissipation rate with respect to FVM. The results of the acoustic pulse simulation on a sufficiently coarse mesh with the piecewise-linear approximation show a good agreement with the analytical solution in contrast to the FVM numerical solution. For the Sod problem, the results of the discontinuities propagation illustrate the dependence of the scheme resolution on the numerical fluxes, troubled cell indicator and the limitation technique choice. The possibility to resolve strong shocks is demonstrated with the Sedov cylindrical explosion test. © Published under licence by IOP Publishing Ltd.

关键词： Finite volume method

来源：评论

学校读者我要写书评

暂无评论

Searching by Code: a New SearchBySnippet Dataset and SnippeR Retrieval Model for Searching by Code Snippets

arXiv

引用

arXiv 2023年

作者： Sedykh, Ivan Abulkhanov, Dmitry Sorokin, Nikita Nikolenko, Sergey Malykh, Valentin Huawei Noah's Ark lab Russia St. Petersburg Department of the Steklov Institute of Mathematics Russia Ivannikov Institute for System Programming Russia

Code search is an important and well-studied task, but it usually means searching for code by a text query. We argue that using a code snippet (and possibly an error traceback) as a query while looking for bugfixing instructions and code samples is a natural use case not covered by prior art. Moreover, existing datasets use code comments rather than full-text descriptions as text, making them unsuitable for this use case. We present a new SearchBySnippet dataset implementing the search-by-code use case based on StackOverflow data;we show that on SearchBySnippet, existing architectures fall short of a simple BM25 baseline even after fine-tuning. We present a new single encoder model SnippeR that outperforms several strong baselines on SearchBySnippet with a result of 0.451 Recall@10;we propose the SearchBySnippet dataset and SnippeR as a new important benchmark for code search evaluation. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

A programming environment for heterogenous distributed memory machines

A programming environment for heterogenous distributed memor...

引用

Heterogeneous Computing Workshop

作者： D. Arapov A. Kalinov A. Lastovetsky I. Ledovskih T. Lewis Institute for System Programming Russian Academy of Science Moscow Russia Code CS Naval Postgraduate School Monterrey CA USA

mpC is a medium level programming language for distributed memory machines (DMM). The language is an ANSI C superset based on the notion of a network comprising virtual processors of different types and performances connected with links of different bandwidths. It allows the user to describe a network topology, create and discard networks, and distribute data and computations over the networks. In other words, the user can specify (dynamically) the topology of his application, and the mpC programming environment will use this (topological) information in run time to ensure the efficient execution of the application on any particular DMM. The paper outlines the principal features of mpC and its programming environment which make them suitable tools to write efficient and portable parallel programs for heterogenous DMM.

关键词： programming environments Bandwidth Workstations Computer networks Debugging User interfaces Network topology Detectors Parallel machines Computer languages

来源：评论

学校读者我要写书评

暂无评论

Deriving FSM-based tests using $a,b-\text{faults}$ for Logic Circuits

Deriving FSM-based tests using $a,b-\text{faults}$ for Logic...

引用

IEEE Computer Society Annual Symposium on VLSI

作者： Andrey Laputenko Nina Yevtushenko Valentina Andreeva Anzhela Matrosova National Research Tomsk State University Tomsk Russia Ivannikov Institute for System Programming of the Russian Academy of Sciences Moscow Russia

ISBN: (数字)9781665466059

ISBN: (纸本)9781665466066

Finite state machines (FSM) are widely used in testing and verification of discrete event and hybrid systems. In order to guarantee fault coverage, various fault models and formal methods are utilized ranging from classical finite state machines (FSM) to logic circuits. The main problem for deriving an FSM-based test suite is the size of an FSM for real digital circuits, and thus, test engineers prefer to randomly derive a prefix of a test sequence and later on extend it with tests for hard detectable faults of a corresponding logic circuit. Usually at the latter step test sequences detecting hard detectable stuck-at faults are considered. In this paper, we are concerned about single functional faults, namely, $\boldsymbol{a},\boldsymbol{b}-\mathbf{faults}$ , of a system of Boolean functions of the combinational part of a sequential circuit such as deleting a variable or a cube in a SoP which represents the ON-set or the OFF-set of a corresponding Boolean function and experimentally evaluate fault coverage of a corresponding complete test suite with respect to fault coverage of tests returned by other methods. The experimental results show high fault coverage of test suites complete with respect to $\boldsymbol{a},\boldsymbol{b}-\mathbf{faults}$ and thus, it can be recommended to utilize such a test suite instead of a complete FSM-based test suite or to extend randomly generated test sequences for detecting $\boldsymbol{a},\boldsymbol{b}-\mathbf{faults}$ not killed with the prefix.

关键词： Boolean functions Latches Automata Very large scale integration Libraries Sequential circuits Behavioral sciences

来源：评论

学校读者我要写书评

暂无评论

Improving Accuracy and Completeness of Source Code Static Taint Analysis

Improving Accuracy and Completeness of Source Code Static Ta...

引用

Ivannikov ISPRAS Open Conference (ISPRAS)

作者： N. V. Shimchik V. N. Ignatyev A. A. Belevantsev Ivannikov Institute for System Programming of the Russian Academy of Sciences (ISP RAS) Lomonosov Moscow State University Moscow Russian Federation

ISBN: (纸本)9781665423311

Static analysis is a general name for various methods of program examination without actually executing it. In particular, it is widely used to discover errors and vulnerabilities in software. Taint analysis usually denotes the process of checking the flow of user-provided data in the program in order to find potential vulnerabilities. It can be performed either statically or dynamically. In the paper we evaluate several improvements for the static taint analyzer Irbis [1], which is based on a special case of interprocedural graph reachability problem - the so-called IFDS problem, originally proposed by Reps et al. [2]. The analyzer is currently being developed at the Ivannikov institute for system programming of the Russian Academy of Sciences (ISP RAS). The evaluation is based on several real projects with known vulnerabilities and a subset of the Juliet Test Suite for C/C++ [3]. The chosen subset consists of more than 5 thousand tests for 11 different CWEs.

关键词： Codes Static analysis programming

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：