“Systems and Means of Informatics” scientific journal

Volume 32, Issue 4, 2022

Content   Abstract and Keywords   About Authors

SELF-TIMED PIPELINE'S SOFT ERROR TOLERANCE ANALYSIS

• I. A. Sokolov
• Yu. A. Stepchenkov
• Yu. G. Diachenko
• N. V. Morozov
• D. Yu. Stepchenkov
• D. Yu. Diachenko

Abstract: Practical self-timed (ST) circuits are implemented as a pipeline, similar to synchronous circuits. Self-timed circuits have a number of advantages in comparison with synchronous counterparts but are redundant in hardware.
The article analyzes the tolerance of the ST pipeline to single soft errors, taking into account its hardware redundancy and assuming that each soft error affects only one circuit's logical cell. Due to their two-phase work discipline and the mandatory indication of the successful completion of the switching in each phase, the ST circuits can detect a soft error and suspend the operation of the circuit until it disappears. A failure-tolerant hysteretic latch as a part of the pipeline stage register bit ensures that the register is immune to any soft error in the pipeline stage's combinational part. The DICE-like implementation of this latch increases the ST register tolerance to internal soft errors by a factor of 2.7. In general, the ST pipeline is 2.5-6.8 times more immune to single soft errors than its synchronous counterpart.

Keywords: self-timed circuits; pipeline; soft error; failure tolerance; indication; hysteretic trigger

DETECTION OF DISTRIBUTION DRIFT

• A. A. Grusho
• N. A. Grusho
• M. I. Zabezhailo
• D. V. Smirnov
• E. E. Tmonina
• S. Ya. Shorgin

Abstract: Changing the properties of the data being collected is often referred to as data drift (various options for shifting the characteristics of the data).
The existence of drift in artificial intelligence system training data often leads to a decrease in the efficiency of machine learning and erroneous solutions of artificial intelligence systems built on these data. In this regard, the problems of detecting drift in machine learning data, the moment of drift formation, and the consequences of changes in training data become relevant. The work proposes a method for detecting the drift of a probability distribution in an arbitrary metric space of large dimension. The method relies on the difference between unknown probability distributions in different regions of the original space in the event of drift. A drift model consisting of two different probability distributions is considered. Using the balls in metric space as the basis of the method allows one to create an efficient algorithm for calculating the ownership of data points to one of the balls associated with different distributions of the drift model.
This circumstance seems to be essential for revealing the drift of a distribution in a high-dimensional space.

Keywords: distribution drift; mathematical statistics; efficiently calculated algorithm

CONVERGENCE RATE AND STABILITY ESTIMATES FOR A CLASS OF NONSTATIONARY MARKOV MODELS OF QUEUES WITH IMPATIENT CUSTOMERS

• I. A. Kovalev
• Y. A. Satin
• A. I. Zeifman

Abstract: A nonstationary queuing system with S servers and impatient customers is considered, i. e., the arrival intensities decrease with the growth of the queue. The process X (t) describing the number of customers in such a system is considered, the existence of a limiting mode of the probability distribution of states and a limiting mean for X (t) is proved, and the estimates of the rate of convergence to the limiting mode and the limiting mean are obtained. Also, the perturbation estimates are obtained. The authors apply an approach based on the concept of the logarithmic norm of the operator function. As an example, a simple model of a nonstationary system is considered in which potential customers are discouraged by queue length.

Keywords: rate of convergence; ergodicity bounds; logarithmic norm; perturbation; queuing systems

METHODS FOR RETRIEVAL OF IMPLICIT LOGICAL-SEMANTIC RELATIONS FROM PARALLEL TEXTS

• A. A. Goncharov

Abstract: The paper focuses on the problem of targeted retrieval of fragments with implicit logical-semantic relations (LSRs) from parallel texts. In order to answer the question whether such retrieval is possible, the author considers the potential of parallel text as a source of new linguistic knowledge and reviews the works that report results of implicit LSRs studies. The aim of the survey is to identify the methods used to create arrays of text fragments with implicit LSRs. It is revealed that parallel texts offer extensive opportunities for retrieval of such fragments, owing to the fact that an explicit LSR in language A may signal an implicit LSR in language B (that is called a bilingual explicit/implicit mismatch). The following methods have been successfully applied for detection of such mismatches: (i) discourse parsing; (ii) reverse machine translation; and (iii) bilingual exclusion search. Therefore, it can be argued that parallel texts - unlike the monolingual ones - can solve the task of targeted retrieval of fragments with implicit LSRs.

Keywords: implicitness; implicit logical-semantic relations; targeted retrieval; natural language processing; knowledge extraction from texts; parallel texts

NOISY TEXT ANALYTICS

• M. P. Krivenko

Abstract: The article is devoted to an overview of methods for interpreting noisy text data in order to obtain significant information from them. Analytics allows one to isolate useful concepts, draw conclusions from the collected data, and form a forecast. It is assumed that the texts being processed may not correspond to the target and selected reference language. Such deviations can be caused by measurement and fixation errors, be the result of the influence of random or unforeseen factors, or arise as a result of incorrect choice or tuning of the model.
The article lists the types of distortions. The areas of application of methods of intellectual text processing are considered: scientific publications; blogging; e-mails; social media; speech messages; and web analytics. The methods focused on the processing of noisy texts are indicated. Promising directions for further research are formulated: clarification of the concepts of "noise" and "dirty" texts; development of ways to measure the degree of anomaly of the text; systematization of analytical tasks of text processing; and formation of criteria for the effectiveness of methods of intellectual analysis of the text to facilitate the selection of suitable technologies.

Keywords: text mining; noisy text; dirty text; analytics; review

TOKENIZATION BASED ON THE METHOD OF FUNCTIONAL PATTERNS

• Yu. V. Nikitin
• A. A. Khoroshilov
• A. E. Makarova

Abstract: The article proposes a new method of text tokenization based on the use of generalized functional templates. The method is based on the classification of Unicode characters in terms of their role in the formation of text elements and on the use of compound patterns from the generalized character classes. Widespread regular expressions are not used here. A specific feature of the method is the use of a sequence of characters as a part of the interval template. The strengths of the method include successful tokenization of complex information objects (numbers, geographic coordinates, names of articles of engineering products, etc.), obtaining the detailed classification of tokens at the stage of their formation, the ability to turn on and off tokenization of a certain type of tokens, as well as adding new templates according to the sample text for additional training of the system.

Keywords: tokenization; segmentation; graphematic analysis; computational linguistics; patterns; substitution; token

THE METHOD OF DATA EXCHANGE MANAGEMENT IN AUTOMATED INFORMATION SYSTEMS WITH SEMANTIC ANALYSIS OF TRANSMITTED INFORMATION ELEMENTS

• M. M. Gershkovich
• T. K. Biryukova

Abstract: The article proposes a method for managing the transmission of data elements between nodes of a geographically distributed automated information system (AIS) aimed at the fastest delivery of the most valuable information to the user in situations where a queue of data packets appears in the communication channel. The concept of an information object (IO) is defined as a set of characteristics of some entity. The concept of a logical data block is introduced as a combination of logically related characteristics of an IO or event. The composition of the logical block can be set by an expert depending on the content of the data in the AIS and on the tasks solved by users. A method for assigning priorities to logical data blocks during transmission over communication channels depending on the semantic content is described. An algorithm for quantifying the value of a logical block and changing the value of information over time is presented.

Keywords: automated information system; information-analytical system; communication channel; information exchange; data priority; data package

THEORETICAL FOUNDATIONS OF DIGITAL EDUCATION: SUBJECT DOMAIN MEDIA OF INFORMATICS AS THE BASE OF ITS OBJECTS' CLASSIFICATION

• I. M. Zatsman

Abstract: The strategy of European digital education (EDE) adopted a dyadic approach to teaching, namely, on the one hand, the study of informatics as an independent discipline, and, on the other hand, the study of information technology, methods, and means of informatics used in other branches of science.
The results of a two-year monitoring of the state of informatics teaching systems in the regions and countries of Europe, including the Russian Federation, as well as in Israel, were used in the formation of the strategy of EDE. The theoretical foundations of the EDE strategy and informatics are fragmentally considered in a report containing the results of the monitoring. The description of the subject domain of informatics is partially given in papers, including the works of Nygaard, Denning, and Rosenbloom, cited in the report. In the subject domain of informatics, Nygaard distinguishes entities of mental nature (concepts of human knowledge) and sensory-perceived entities (for example, texts in natural languages). The paper proposes to separate out media from the subject domain of informatics based on the approaches of Nygaard, Denning, and Rosenbloom.

Keywords: education; dyadic approach; informatics as a fundamental science; informatics as an educational discipline; strategy of teaching informatics

NATIONAL RESEARCH AND EDUCATION NETWORKS OF THE EURASIAN ECONOMIC UNION COUNTRIES: CURRENT STATUS AND PERSPECTIVES OF INTEGRATION

• A. G. Abramov
• A. A. Gonchar
• A. V. Evseev
• A. T. Idrisov
• D. V. Novik
• B. M. Shabanov

Abstract: The topic of the paper refers to the systematization and discussion of the prevailing prerequisites and prospects for the creation of an association of national research and education networks of the Eurasian Economic Union (EAEU) countries. Some aspects of the digital transformation of the sphere of science and education and the role of sectoral telecommunications networks in it are touched upon. The current characteristics of the national research and education networks of the EAEU countries, the key components of the developed concept of creating, ensuring the functioning and development of the network association are given. The proposed technological solutions for the formation of the telecommunications infrastructure are discussed; a forecast of the expected effects from the creation of the association and the further development of the project is given.

Keywords:  National Research and Education Network; NREN; EAEU; National Research Computer Network; NIKS; KazRENA; telecommunications infrastructure; association of research and education networks

SOME PROJECTS OF NONSTANDARD EXPERT SYSTEMS APPLICATION

• D. V. Zhukov

Abstract: The issues of specialized information systems functionality developing through the introduction of intelligent solutions based on artificial intelligence technologies are discussed. The proposed solutions are based on the results of a comparative analysis of the implementing state strategies processes in the field of artificial intelligence and the specific conditions in the Russian Federation. Four perspective projects of expert systems are presented. Along with the "standard" solution based on the facial biometrics technology, three "nonstandard" projects rely on the classic first-wave artificial intelligence solutions which have not received as much attention as projects based on the second-wave artificial neural networks.

Keywords: artificial intelligence technologies; expert systems; artificial neural networks; performance management; provision of resources

CLUSTERING ALGORITHMS FOR TECHNOLOGY OF CONCRETE HISTORICAL INVESTIGATION SUPPORT

• I. M. Adamovich
• O. I. Volkov

Abstract: The article continues the series of works devoted to the technology of concrete historical research support. The technology is based on the principles of co-creation and crowdsourcing and is designed for a wide range of users which are not professional historians and biographers. The article is devoted to the further development of the technology by integrating into it a mechanism for automated search for anomalies in concrete historical information based on cluster analysis.
The analysis of the specifics of concrete historical data and the ways of their representation in the object model of technology is carried out. The methods of mixed data digitizing and the proximity measures used for them are considered in detail and the advantages and disadvantages of clustering algorithms used to search for anomalies are evaluated. Based on the analysis, an approach was developed to search for anomalies in the data of technology and directions were outlined for testing the effectiveness of the selected algorithms and proximity measures on real concrete historical data.

Keywords: concrete historical investigation; distributed technology; anomaly; historical-biographical fact; clustering

METHOD OF TARGET DISPLACEMENT OF SOLUTION IN SITUATIONAL MANAGEMENT TECHNOLOGIES

• A. V. Ilyin
• V. D. Ilyin

Abstract: The updated technology for interactive solving of linear problems of situational management by the method of target displacement of solution is considered. The formulations of tasks and the requirements for the quality of solutions are based on the analysis of portraits of the target and achieved situations. The method of target displacement of solution (as opposed to the standard linear programming methods) allows the expert to get the desired result even if the constraints system is incompatible. The interactivity of the method allows the expert to change the problem formulation at each step of the solution search based on his/her opinion on the feasibility and effectiveness. The technology is focused on implementation in the form of online service.

Keywords: situational management; linear problems of situational management; resources allocation; portrait of the situation; method of target displacement of solution; online service

OPTIMIZATION PROCESS OF THE INTEGRATED PLANT PROTECTION METHOD FOR DOT MODELS

• R. N. Odinaev
• P. L. Nazrulloev
• F. Raimzoda

Abstract: An integrated method of protection against pests of agrocenoses is understood as the complex application of chemical, biological, agrotechnical, mechanical, and other methods of protection against harmful insects. The parameters of the integrated protection method are the level of efficiency of entomophages and the level of harmfulness of pests which are presented in the form of mathematical formulas. The present article discusses the optimal control problems associated with biological systems consisting of three trophic levels "plant - harmful insects - beneficial insects" of the plant protection process taking into account the incoming external resource. For a model biosystem, the necessary conditions for solving the optimization process of plant protection from pests have been formed and substantiated. The Pontryagin maximum principle is proved for optimal control problems associated with model biosystems.

Keywords: optimal control; agrocenosis; beneficial insects; harmful insects; integrated method; plant protection; insect abundance; cost functional; dot model

SEARCH OPTIMIZATION WHILE SOLVING ENUMERATION PROBLEMS IN AN ADVANCED COMPUTER SCIENCE COURSE AT THE LEVEL OF BASIC GENERAL EDUCATION

• O. M. Korchazhkina

Abstract: The article discusses a few ways to activate high school students' logical thinking to form their algorithmic skills in computer studies lessons.
The research analyses the methods for optimizing in traditional cryptoarithmetic problems that make up a subclass of enumeration tasks which were proposed by Herbert Simon to illustrate the ways of mindset when it is necessary to carry out numerous iterations of possible combinations. A search optimization visual method is proposed for solving two Simon problems based on the patterns of digit transfer during summation and presented in the form of simple algebraic relations. The method can be easily algorithmized and implemented with dynamic programming technology to optimize the solution of iterative tasks.

Keywords: optimization of a search; enumeration/iterative problem/task; crypto-arithmetic problem; dynamic programming