Software & Systems

Articles of journal № 4 at 2017 year.

Order result by: Public date | Title | Authors

1. A software package for nonuniform time series analysis based on continuous wavelet transformation [№4 за 2017 год]
Authors: Prokhorov S.A., A.A. Stolbova
Visitors: 7417
Wavelet transformation is one of the methods of data time-frequency analysis. In practice, the researcher often has to analyse nonequidistant (uneven) time series. For this reason, the paper considers algorithms for their continuous wavelet transformation. The authors propose an algorithm for obtaining an array of transformation even shifts, taking into account uneven source date. It means that they choose the interval of involuntary discretization, determine the number of shifts and then calculate their value. The proposed algorithm is the base for an algorithm of continuous wavelet transformation of nonequidistant time series. The process of estimating transformation coefficients uses only samples of time series that are in the width of a wavelet. The advantage of thу proposed algorithm is that the result of the transformation is an even representation. The developed algorithms are used in a complex of programs. The paper shows the results of experiments for models of nonuniform time series with p-transformation and with “jitter” (with uniform distribution and triangular distribution). The velocity of the algorithm can be increased by taking into account the effective radius of the mother wavelet and calculating its width.

2. Using concept maps for rule-based knowledge bases engineering [№4 за 2017 год]
Authors: Dorodnykh N.O., Yurin A.Yu.
Visitors: 11834
Using conceptual models in the form of concept maps for engineering rule-based knowledge bases of intelligent systems remains relevant. This relevance demands the development of specialized algorithmic and software. This paper considers an approach to prototyping of rule-based knowledge bases of expert systems based on analysis of IHMC CmapTools concept maps. The approach is based on the extracting structural elements of concept maps from the CXL files (Concept Mapping Extensible Language) and their transformation to the elements of a programming language, in particular, the C Language Production System (CLIPS). The paper describes the main stages of the approach, analyzed constructions of CXL files (in particular, concept-list, linking-phrase-list, connection-list). It also presents an illustrative example of transformations. A distinctive feature of the proposed approach is using an ontological model as a universal intermediate form of knowledge rep-resentation derived from concept maps, which is independent to the knowledge base programming language. Another feature is the author’s graphic notation – Rule Visual Modeling Language (RVML) that provides visualization and modification of cause-effect re-lations as logical rules. The considered algorithms are implemented as a part of a software research prototype called the Personal Knowledge Based De-signer (PKBD). Currently, it is used in the educational process at the Irkutsk National Research Technical University (INRTU) in “CASE-tools” and “Sowtware tools of information systems” courses.

3. Estimation of website security status as badly formalizable objects based on fuzzy logic methods [№4 за 2017 год]
Authors: D.T. Dim, Bogatikov V.N., Klyushin A.Yu.
Visitors: 8387
The article analyzes page load time, which is an important indicator for any website. Websites are typically hosted on web servers with certain characteristics. These websites interact with the environment, which is mostly aggressive and undefined (external threats, such as penetration, service denial, SQL code injection, etc.). It should be noted that there are uncertainties caused by hardware and software, which arise due to the reliability properties of software and hardware products. Thus, any website can be influenced by external factors and various subsystems serving website functions, which leads to abnormal situations and website operation uncertainty. The degree of uncertainty is not always practically feasible to estimate only based on statistical material. This leads to increasing the number of methods and means for intellectualizing the performance of estimations based on artificial intelligence methods, particularly fuzzy estimate methods.

4. Analysis of website security status based on performance metrics [№4 за 2017 год]
Authors: D.T. Dim, Bogatikov V.N.
Visitors: 4296
The paper is relevant due to the constant evolution of Internet infrastructure including conditions for optimal operation of web systems, information sources, e-commerce and various types of services. During the past twenty years, general information in the World Wide Web (WWW) and the number of users has increaseв almost two hundredfold. With this growth, requirements for technology has become even more critical. However, current research is more about a commercial aspect then the technical one. The paper pays special attention to assessing website performance quality regarding information technology indicators. It also considers a fuzzy Markov chain to define specific website productivity states.

5. Optimum entropy clustering in information systems [№4 за 2017 год]
Author: B.G. Askerova
Visitors: 8491
The paper researches the possibility of developing a new method for data clustering in information systems. Clustering is a process of searching possible groups in a given set using signs of similarity or difference between elements of this set. The existing entropy clustering method includes an information theoretic approach to a clustering task. The paper suggests a clustering method based on an entropy approach to selecting message items. The paper suggests a method of optimum entropy clustering of high-dimensional data in information systems. It also gives mathematical grounding of the method. The suggested method of optimum entropy clustering is based on the known principle “low entropy corresponds to big information content”. This make it possible to form an optimum clustering regime, as well as an attribute space reduction regime. The paper proposes a method for calculating a level of clustering optimality. It also describes a method for reducting attribute space of high-dimensional data upon their initial processing.

6. Architecture of an intelligent optimal control system for multi-stage processes evolution in a fuzzy dynamic environment [№4 за 2017 год]
Authors: Palyukh B.V., Vetrov A.N., I.A. Egereva
Visitors: 9386
The article considers key elements of the approach to creating a system of intellectual information support of innovations at the enterprise. The approach is based on the integration of search mechanisms for innovative solutions and methods of industrial-technological system evolution control using the created store of innovative solutions, algorithms of coordinated optimization and process parameters identification. The paper examines the possibility of using the proposed approach with respect to the base version of the industrial-technological system functioning model. The project is dedicated to a fundamental scientific problem of developing methods and model means of optimal control of multi-step processes evolution under conditions of dynamic uncertainty to enhance their effectiveness and long-term sustainability throughout their life cycle. The authors consider a multistage process as a multi-agent system, its effective management depends on consistent behavior of the center and agents, their interest in finding and implementing innovative solutions, ability to analyze the possibilities of evolutionary development. The approach behind the proposed in the project of a formal apparatus for optimal control of multi-step process evolution includes development and study of a mathematical model of managing multi-stage processes evolution in a fuzzy dynamic environment. It also involves development of a method of solving the problem of managing multi-stage processes evolution in the form of the optimal (suboptimal) control law with feedback and the study of asymptotic properties of solutions for a functional equation obtained for autonomous systems. The project content also includes development of methods for coordinated optimization when the center interacts with agent groups in terms of expanding their production and technological capabilities. Implementation of the developed methods and models involves creating a prototype of an intelligent system for optimum control of a multi-stage processes evolution in a fuzzy dynamic environment.

7. Cognitive hybrid systems for decision support and forecasting [№4 за 2017 год]
Authors: A.N. Averkin, Yarushev S.A., V.Yu. Pavlov
Visitors: 17626
The paper considers a number of models for decision-making support in dynamic situations that are characterized by weak structuring based on a hybrid system integrating a fuzzy hierarchical assessment model and a fuzzy cognitive situation model. The paper presents a hybrid model based on cognitive maps and Saati decision support hierarchies in dynamic situations and a fuzzy production model for modeling people's irrational behavior in the problems of behavioral economics. When creating a behavioral decision-making model, we were taking into account the modules responsible for decision maker’s emotions and internal model representativeness. The model uses fuzzy logic and production rules. This approach makes the decision-making model intuitive due to linguistic variables that form production rules. Another advantage is universality and scalability obtained when switching to models with a large number of parameters. The paper presents a model of a modular time series forecasting system. It consists of modules based on modular neural networks, a module that includes a fuzzy cognitive map hybrid and a neural-fuzzy ANFIS network and modules for verifying and aggregating the results. This article considers in detail a module combining a fuzzy cognitive map and a neural-fuzzy network. There is the constructed neural network and its structure is shown in combination with a fuzzy cognitive map based on the forecast of the indicator “living standards”. Business analytics systems use similar approaches for knowledge economics based on intelligent decision support systems that use cognitive methods of analyzing consciousness of people involved in these processes, as well as testing decision makers’ quality by their brain activity, for parametric adjustment of intelligent systems decision support.

8. Classification and clustering methods for improving efficiency of case-based systems [№4 за 2017 год]
Authors: Varshavskiy P.R., Ar Kar Myo, D.V. Shunkevich
Visitors: 12994
The article examines topical issues of improving efficiency of case-based reasoning (CBR) systems. Case-based methods and systems (CBR systems) are actively used to solve a number of problems in the field of artificial intelligence (for example, for modeling plausible reasoning (common sense reasoning), machine learning, intellectual decision support, intelligent information search, data mining (DM) and etc.). It should be noted that modern tools for DM, which are widely used in intelligent systems, database and knowledgebase management systems, business ap-plications, machine learning systems, electronic document management systems, etc., do not have advanced CBR tools. The paper proposes to use the modified CBR cycle to increase the efficiency of CBR systems. This cycle allows creating a base of successful (CB) and unsuccessful cases (UCB) based on available expert information (test samples), as well as the k-nearest neighbors (k-NN) modification algorithm for case retrieval. The proposed modifications allow improving the quality of solving DM tasks (in particular, data classification task). In addition, the authors consider the possibility of reducing the number of cases in CB using classification and clustering methods to improve performance of CBR systems. The paper shows computational experiments to estimate the effectiveness of the solutions offered while working on a data set from the UCI Machine Learning Repository. They use CBR system prototype developed in MS Visual Studio in C# language.

9. Fuzzy data storing and efficient processing in PostgreSQL DBMS [№4 за 2017 год]
Author: Sorokin V.E.
Visitors: 9896
In just half a century the theory of fuzzy sets has developed from theoretical foundations to practical use in artificial intelligence systems in such areas as industrial equipment and vehicles control, medical diagnosis and expert systems, including risk assessment, from economic to environmental. Many of them need efficient and reliable storage and processing of large volumes of information. For this purpose, there are DBMS, the most advanced of them are the object-relational DBMS. However, fuzzy data are difficult to reconcile with both object and relational data models. Moreover, the majority of industrial DBMS do not contain built-in fuzzy data types and mechanisms to work with them. These DBMS include cross-platform and freely distributed in source object-relational DBMS called PostgreSQL. Import substitution in infrastructure software is largely associated with application of PostgreSQL. Opportunities of fuzzy data relational modeling and a developed mechanism for extending types through the creation of the required user-defined types along with powerful procedural languages in PostgreSQL allow implementing various alternative approaches to fuzzy data storage and handling. The article shows a comparative analysis of these approaches from the point of view of maintaining data integrity and processing efficiency. There are the experimental results at various fuzzy data models depending on approaches. The paper proposes design solutions that increase the efficiency of fuzzy data search, and formulates recommendations on choosing an approach to modeling fuzzy data.

10. Interpretation of the meaning of natural language phrases in problem-oriented systems [№4 за 2017 год]
Authors: Billig V.A., I.S. Smirnov
Visitors: 5601
The organization of a natural language dialog with a computer is one of the most important problems in the field of artificial intelligence. The expressive power of a natural language makes it difficult to formalize and eliminate ambi-guities in understanding phrase meanings. This article considers the approach to interpreting natural language phrases based on the “Meaning–Text” theory. The key point is an intentional dialogue context limitation with a specified domain. This allows the system to conduct a more meaningful dialogue and to solve specific problems of the given domain. The “Eliza–Student” software built upon the developed algorithms is oriented to a fairly broad subject domain, which includes a fair amount of tasks from the Unified State Exam (the Russian abbreviation is “ЕГЭ”, an English analogue is SAT) in informatics. The developed system is able to explain actions performed during the analysis of the text, which contains user questions on solving certain tasks in informatics, solve them and explain decision making in solving process. It seems that the reciprocal form of interaction seems to be the most natural in the learning process. The analysis process is divided into a sequence of stages (preliminary, morphological, syntactic and semantic analysis). Each of them uses different models of language and subject domain. The proposed approach is based on the following ideas: abstraction from the subject area to the latest stages of analysis and focus on the result, i.e. the construction of the most probable, perhaps incomplete, representation of the meaning, despite the incompleteness of the initial information or possible errors in the analysis process. The developed algorithms can be applied to different subject domains.

| 1 | 2 | 3 | 4 | Next → ►