ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 12. Vol. 22. 2016

To the contents

A. A. Razorenov, Ph. D. student, e-mail: arazorenov@hse.ru, V. A. Fomichov, Professor of Computer Science, e-mail: vfomichov@hse.ru, National Research University Higher School of Economics (HSE), Moscow, Russia

Compact Formalization of Instructions Semantic Parsing Algorithms' Input and Intermediary Data Structures

The paper suggests new formal means for the development of the algorithms fulfilling semantic processing of natural language instructions. The theory of K-representations (knowledge representations) is chosen as the methodological basis of the study. Nowadays it is the only theory being convenient for constructing semantic representations of arbitrary complex natural language instructions. The following theoretical results are set forth: (a) a new formal model of a linguistic database is developed, this model is considerably more compact than two models introduced by the theory of K-representations; (b) the notion of semantic-syntactic representation of a natural language text is formalized in a new way. The developed formal means are used in the design of a useful for practice natural language interface of file manager.
Keywords: natural language processing, semantic parsing of instructions, semantic representation, theory of K-representations, SK-language, morphological basis, template of a K-representation, dictionary of lexical frames, linguistic database, natural language control of file system, Natural Language Commander

P. 883–891

A. S. Domnikov, Associate Professor, V. V. Belous, Associate Professor, walentina.belous@bmstu.ru Moscow State Technical University n.a. N. E. Bauman, Moscow, Russia

Synthesis of Rational Taught Modules Ordering by Berzha's Method

The work considers a problem of creating a rationally ordered electronic course modules. Features of this task are investigated and the conclusion is drawn that it is impossible to solve it by classical methods of mathematical programming and mathematical statistics. The expert poll or data included in hypertext structure of links provides initial information for the synthesis of a rational order of the taught modules. The paper considers two types of expert information: clear and fuzzy binary relations of preference based on a set of the taught modules. It offers to reduce all three structures to the normalized view, i.e. a matrix of paired comparisons with simple calibration. The proposal is to solve a problem of ordering by Berzha's method, which has received wide practical approval in a diversity of applied fields.
Keywords: electronic course, ordering, theory of decision-making, binary

P. 892–898

I. G. Malygin, Professor, Director, V. I. Komashinskiy, Professor, Deputy Director, kama54@rambler.ru
IPT RAS

Information Technology and Artificial Intelligence —the Main Engines of the Fourth Industrial Revolution (Industrie 4.0)

In paper the fundamental features of new industrial revolution in motor transport sphere are observed. It is shown, that a key technological platform for new industrial revolution (it is not dependent on spheres of its application) are êîãíèòèâíûå informational-network production engineering integrated (more precisely, converged) with production engineering of an industrial artificial intellect. It is marked also, that the important condition of successful conducting of new industrial revolution is preparation of professionals, capable to develop and establish intellectual industrial infrastructures and to sustain their effective operation and the subsequent development.
Keywords: fourth industrial revolution, industrial information and control networks, industrial artificial intelligence, cognitive industrial systems, intelligent transport

P. 899–904

I. Ya. Lvovich, Professor,  Paneuropean university, Bratislava Ya. E. Lvovich, Professor, office@vivt.ru, A. P. Preobrazhensky, Professor, app@vivt.ru, O. N. Choporov, Professor, choporov_oleg@mail.ru
Voronezh institute of high technologies, Voronezh

The Features of the Methods of Computational Fluid Dynamics for Turbulence Modeling

In the paper an analytical review of methods for turbulence modeling on personal computers is carried out. Turbulent flow can be represented as a multiple rotating in different directions and in different planes of vortices. The computational features of the methods that can be used in mathematical software subsystems for the simulation of turbulence in the relevant CAD are provided. The results that computational fluid dynamics can take a dominant place, surpassing the prevalence of experimental and analytical methods are shown. These differences in the three main methods of description of turbulence are discissed. It was provided that, the development of computer technology and numerical methods of computational physics, allows to work not only with increasingly complex models, but also to apply different ways of visualizing the obtained result. The obtained results can be helpful when selecting a turbulence model for a specific task in the simulation subsystems.
Keywords: turbulence, modeling, software, CAD, computational physics, equation, vortex, integral engineering application, liquid,
method

P. 905–913

E. M. Bronstein, Sc. D., Professor, e-mail: bro-efim@yandex.ru, A. A. Davletbaev, Postgraduate Student, e-mail: Arvid_NF@mail.ru, Ufa State Aviation Technical University

Vehicle Routing Problem with Transport Cost Depending on Weather

The problem of constructing a cyclic route with minimal transport cost for delivering a homogeneous cargo from a production point to consumers by limited capacity vehicle is considered. The cost of transportation between points depends on time. The corresponding integer linear and quadratic models are constructed. Computer experiments solving problems of different dimensions and comparative analysis efficiency of models were carried out in environment IBM ILOG CPLEX Optimization studio 12.6.1.
Keywords: routing, nonstationarity, linear integer programming, branchs and cuts method

P. 913–917

V. B. Yeflov, Associate Professor, e-mail: veflov@gmail.com,
E. A. Pitukhin, Professor, e-mail: eugene@petrsu.ru, Petrozavodsk State University, F. V. Stafeev, Consultant, e-mail: fedor.stafeev@gmail.com  Petrozavodsk City Administration

Performance Issues of Query Execution in Relational H2 Database

To exercise the powers of modern local authorities a wide range of ICT tools is used. A key component is the use of DBMS in various information systems. Despite the long-term development of optimizers of DBMS, in industrial information systems operation an issue of improving the performance of the query modules is becoming quite topical.
To achieve this goal it is necessary to solve the problem of development of performance assessment methods of the query execution modules of a freely distributable relational DBMS H2 in case of varying parameters of hardware and software platforms and predicates of SQL expressions.
The statistical characteristics of the query execution time have been studied; the features of the work of the Cached exchanged mode have been identified. Using the nonlinear approximation of the experimental data the generalized model of dependence of the performance of query execution on the amount of cache memory and the size of the requested tables was obtained. A dimensionless parameter of the relative amount of cache as the volume ratio of the cache to the database volume was introduced, based on which a criterion for the implementation of efficient caching mode in DBMS H2 was proposed.
The present study shows a fairly significant impact of caching on the query execution process, especially in the tables of a relatively small volume. The generalized model based on run-time queries of the cache size and the amount of the requested tables, which was obtained as a result of approximation, allows us to conduct forecasting of the query execution time at the determined conditions for any mode of cache operation. The simulation results demonstrate that from the practical perspective it is important to ensure the operation of DBMS H2 with the use of caching feature in the "effective" mode, when the allocated capacity of the cache memory is sufficient for the provision of efficient operation of the system. As a criterion ensuring this mode, it is possible to use exceeding the proposed dimensionless parameter of the relative volume cache against a certain critical level. The achieved results can be used in practice to significantly reduce the query execution time.
Keywords: query optimization, h2, relational database, math modelling, query execution, caching

P. 918–924

E. Yu. Orekhov, Associate Professor, e-mail: emil.orekhov@bk.ru, Yu. V. Orekhov, Associate Professor,
Ufa State Aviation Technical University

Comparing and Estimating Quality of Heuristic Algorithms on a Finite Problem under Non-Equiprobable Generation of Test Instances

Efficiency of heuristic algorithms can be estimated by introducing a "functional" quality characteristic for a given heuristic algorithm on a given finite problem. This characteristic is a distribution function of a random variable that takes on a set of possible values of the algorithm’s quality criterion on problem instances. This fact allows statistical estimation of the quality characteristic by a certain random generation of points within a region of parametric space for the given problem. These points correspond one-to-one to the problem instances. However, numerical quality characteristics are more common in practice, such as the expectation of a random variable. Equiprobable generation of problem instances ensures the appropriate statistical estimate for the quality characteristic, but the task of developing equiprobable generators arises here. In this paper we suggest an approach to comparing and estimating quality of heuristic algorithms based on the expectation of the random variable mentioned above, in cases when equiprobable generation of problem instances is difficult or impossible. This approach involves a secondary set of objects associated to the set of problem instances, while equiprobable generation of elements of the secondary set is quite easy. We give an example of applying our approach to statistical estimation of quality for heuristic algorithms that solve a problem of scheduling jobs between unrelated parallel machines. We also define the conditions of application for the suggested approach.
Keywords: heuristic algorithm, finite problem, problem instance, quality characteristic of a heuristic algorithm, quality criterion of a heuristic algorithm, statistical estimate for the expectation of quality criterion, equiprobable generation, scheduling jobs betweenunrelated parallel machines

P. 925–930

S. V. Gavrilov¹, DSc, Head of Department, S. I. Gurov², PhD, Assistant Professor, T. D. Zhukova¹, Research Engineer, D. I. Ryzhova¹, Junior Research Scientist, ryzhova_d@ippm.ru
¹Institute for Design Problems in Microelectronics of Russian Academy of Sciences (IPPM RAS),
²Lomonosov Moscow State University

Application of Coding Theory to Improve the Noise Immunity of Combinational Circuits

The work is devoted to solving problems of design automation to improve the noise immunity of combinational integrated circuits at the logical level design. The different methods of error-correcting coding for improving noise immunity of combinational integrated circuits are analyzed. The coding methods with a selection of duplicate circuit outputs commutation based on the logical correlations between the circuit outputs are implemented. One of the significant difference of the noise immunity of combinational circuits, in comparison with coding in data networks, is an additional degree of freedom associated with the implementation of the coding scheme. The application of effective optimization methods reduces the size of coding circuit by selection of the commutation order, compared to the majority approach.
Keywords: combinational circuits, design, noise tolerance, CAD

P. 931–937

E. N. Reshetova, Teacher, Software Engineering School, Faculty of Computer Science, HSE, ereshetova@hse.ru

Student Universal Competences of Educational Direction "Software Engineering"

It is considered one of the most pressing problems of education at the present stage — definition of the universal competence of technicians. We present some of the current approaches to this issue and propose a model of universal jurisdiction as mandatory intellectual basis of professional software engineer.
Keywords: software engineering, universal competences, SWECOM

P. 938–945

B. I. Soloviev, employee, D. L. Zhusov, employee, e-mail: d.zhusov@mail.ru, S. À. Prosolupov, employee The Academy of Federal Security Guard Service of the Russian Federation

Determining the Placement of Network Resources on a Quasi-Homogeneous Topological Structure of Information and Communication Systems

The paper presents the formulation and procedure for determining the locations of resources on the network information and communication structure. The aim of this work was the development of the method provides enhancing the stability of user access information and communication systems. The paper presents a mathematical model of the system topology. The formulated goal is achieved by representation of the structure of the info-communication network as a random graph and using step scheme of analysis, design, and elimination properties of heterogeneous resources. A distinctive feature of the proposed solution is solving the dropout unpromising initial parts of options before their full count of the build.
The procedure of calculations in accordance with the method of successive analysis of variants and the proposed scheme on the example of implementation of the requirements on the bandwidth information and communication systems.
The solution of a problem of determination of locations of network resources on the nodes and edges of the invariant part of the topological structure of the information communication system allows you to organize sustainable user access regardless of location on different routing schemes in any proportion.
Keywords: structure, resources, invariant part, two-submultiple count, center of the count, median of the count

P. 946–951

V. N. Tarasov, Head of Chair, Professor, e-mail: tarasov-vn@psuti.ru,
N. F. Bahareva, Head of Chair Professor, L. V. Lipilina, Assistant, Volga State University of Telecommunications and Informatics, Samara

Automation for Calculating Characteristics Queuing System for a Wide Range Changing Their Parameters

The article describes the characteristics of the queuing design software systems (QS) G/G/1, which correspond to the input distribution with a coefficient of variation of the interval between the receipt of the request and service Times lowest and highest 1, that is, cover the entire range (0, ¥). To do this, we consider the QS M/M/1 with delay, H₂/M/1 and H₂/H₂/1 with Hyperexponential input distributions 2nd order. As the output characteristics used by the average waiting time in the queue and its variance.
Other characteristics of the systems in question are derived from the average waiting time.
With the definition of a delay variation (jitter) in a telecommunications time-variation of its mean value, thus having the opportunity to determine the jitter through dispersion. This result is important for the analysis and description of traffic that is sensitive to delays. For receiving output characteristics the method of spectral expansion of the solution of an integral equation of Lindley is used. For the practical application of the results used the traditional method of moments. It is known that distributed according to the law Hyperexponential H2 random variable has a coefficient of variation greater than 1. In [1] it is shown that the coefficient of variation of a random variable distributed according to an exponential law with a time lag of less than 1.
Given the fact that the distribution of H2 is a three-parameter in the software system has a mechanism of approximation, arbitrary laws Hyperexponential distributions at the level of the first two moments, and at the level of the first three moments.
Keywords: queuing systems, the average waiting time in the queue, the method of spectral decomposition, the integral equation Lindley, Laplace transform

P. 952–957

To the contents