Purpose of research. The goal of the work is to develop and justify a cognitive peer-to-peer infrastructure that will improve the conditions for collective work on projects based on agile methodology. Cognitive architecture is defined as a structure that ensures the implementation of anthropomorphic and neuromorphic functions in natural or artificial systems. The proposed approach is based on organizing the interaction of the collective intelligence of members of an agile team and artificial intelligence, represented by trained artificial neural networks. When forming an agile team, it is proposed to take into account the structure of the cognitive sphere in the structure of the mental processes of a human cognitive agent.Methods. Domain knowledge is determined based on the collective intelligence of the agile team members and the training of artificial neural networks. It is assumed that artificial neural networks are available to all members of an agile team and implement the functions of collective artificial intelligence, provided that their training uses the professionalism and experience of a person in a natural social environment. Mental operations such as analysis, partitioning (modularization), comparison, abstraction, synthesis, generalization, classification, concretization, known from general psychology courses, are interpreted not only as a result of human activity, but also as the functionality of a program. Some elements of the cognitive sphere processes “memory” and “speech” are realized in a similar way.Results. The system is implemented on the basis of a peer-to-peer computer network that provides communications between all artificial and natural participants in the cognitive process during the design process. A conceptual model of a cognitive collective intelligence cell is proposed, combining elements of the actual collective intelligence of agile agents with the collective artificial intelligence of agents based on neural networks. In an expert assessment of the quality of individual design stages, it was proposed to use tagging based on the emotional-volitional and motivational mental processes of individuals.Conclusion. Cognitive information processing is based on the idea of modeling human thinking processes in computer systems. In the system under consideration, this includes natural language processing, written speech recognition, associated with understanding information through software imitation of human intelligence. The accepted concept involves the implementation of collective intelligence not only artificially, but also by organizing convenient interaction between participants in an intellectual chat. Artificial intelligence, also collective, is implemented using initially trained and further trained neural networks.

Purpose of research. The purpose of the work is to develop recommendations for the software implementation of cognitive agent-based systems that ensure interoperability in the interaction of software cognitive agents with different properties. A software implementation that determines semantic proximity based on machine learning can automatically and quickly highlight important key concepts and find associations, simplifying and speeding up the process of working with text data during a dialogue between agents, one of which is a human. The proposed approach is based on the assumption that computer systems can perform some “anthropomorphic” functions, similar to human ability to think.Methods. Domain knowledge is determined by training an artificial neural network. To indicate the semantics of remarks and other information, it is proposed to use tagging and determining the semantic proximity of key phrases from speeches presented in written form.Results. The system was implemented in the Python programming language. The Word2Vec model with Skip-gram architecture was used as a neural network model for text vectorization. For training, two text sets with information about computer science and zoology were used. Based on the results of comparing texts on two topics, one can judge the performance of the system to determine the semantic proximity of textual information.Conclusion. The subsystem for determining the semantic proximity of text information based on machine learning technologies, which forms the basis for the software implementation of cognitive interoperable agent-based systems, will improve the efficiency of existing or developed applications that involve a large amount of text information.

Background. Currently, there is an intensive search for solutions in the field of creating infrastructure for network applications, for example, for the decentralized Internet based on several main technologies: blockchain, machine learning, Semantic Web and Internet of Things. The possibility of organizing the infrastructure of the decentralized Internet, or Distributed Web – DWeb, in the form of a peer-to-peer (P2P) network, the nodes of which are user devices, is being considered. The aim of this work is to study one of the approaches to the practical implementation of the application-driven functional architecture of the ADFA-DCS (Application-Driven Functional Architecture of Distributed Computing System), defined by conceptual and logical artificial intelligence models. Materials and methods. A new functional architecture of the DCS is proposed, which is determined by the application, the implementation of which differs from the known ones in that the operation of the application is preliminarily determined by the executable model – the conceptual Petri net graph. Then a structure is formed that specifies the logic of the application, and on its basis a distributed control program is generated. This principle makes it possible to implement arbitrary distributed algorithms without significant costs for reprogramming the application. The authors of this work believe that the concept of a distributed network with the proposed architecture is in good agreement with the platform of geographically distributed corporate networks SD-WAN (Software Defined Wide Area Network), the construction of which implements a fundamentally new approach to network management. Results. The proposed functional architecture is implemented on a global (WAN) or local area network platform (LAN), on which the logical system architecture of a distributed computing system is defined. To implement the system architecture of the DCS under consideration, technologies close to peer-to-peer technology were selected. The process of programming a distributed application is accompanied by the use of message and data transfer functions, searching for data in tables, copies of which are located on all nodes of the P2P network, and the implementation of operators performing functions for which the created network DCS was intended.

Purpose of research. The main purpose of this work is to improve the quality and efficiency of managerial decisionmaking based on the development of a method for assessing and forecasting economic risks of an enterprise. This method is based on data mining technology.Methods. The paper uses methods of panel data processing and analysis, for which a mathematical model for predicting the level of competitiveness of an enterprise was built, as well as a model for predicting economic risks of an enterprise based on combining several methods of data mining: clustering of merging panel data for assessing economic risks of an enterprise and the method of merging fuzzy correlation for statistical analysis of panel data.Results. As a result of the application of the developed method, quantitative assessments of the level of competitiveness and economic risks of the enterprise were obtained. Based on the obtained quantitative assessments of the level of competitiveness and the level of economic risk, a cluster analysis of enterprises in some industry was carried out. The developed methods have high accuracy in predicting economic risks of enterprises, improve the capabilities of data mining and combining information about economic risks of enterprises, which increases the competitiveness of enterprises.Conclusion. A method of forecasting economic risks of an enterprise based on data mining technology has been developed. Weighted estimates of spatial features of panel data were obtained, which allow to obtain integral estimates of the economic risks of the enterprise and the level of competitiveness of the enterprise. A model for the analysis of fuzzy rules of semantic features of panel intelligent data analysis of the assessment of economic risks of the enterprise is proposed. The analysis shows that the developed method has high accuracy and better protection against interference when predicting data.

Background. Based on the proposed methodology for deep structuring of knowledge in a semantically-oriented intelligent computing environment based on expanding the descriptive capabilities of Petri nets by integrating them with conceptual graphs, a technique for interpreting conceptual event network models for given subject areas is given. The computing environment is understood as a virtual distributed computing system implemented on a global computer network. Like the “Semantic Web”, a conceptual graph integrated with a logical Petri net within a single formalism suitable for subsequent machine processing is chosen as the basis for representing intellectual information about a subject area. The schematic representation of conceptual graphs hides much of the complexity associated with predicate calculus. The presented examples of conceptual graphs with event concepts contain not only a declarative, but also a procedural component of the knowledge representation model. The integration of conceptual graphs with logical Petri nets, considered as a number of examples from various subject areas illustrates a special type of semantic networks. The software through which this integration is implemented within the framework of one knowledge representation model is also described. It is shown that subject areas determine the structuring of computer science itself and its directions of development, which also applies to intelligent systems in particular. The purpose of the study is to automate the selection of facts and inference rules for the subsequent software implementation of this approach in intelligent event systems using the example of specific subject areas of human activity. Materials and methods. The methodological basis for researching the subject area is focused on the use of simulation modeling of intelligent event systems, in which the interactions of components are specified locally. In the general case, the construction of a simulation model of an intelligent event system is based on the analysis of cause-and-effect situational relationships and the rules for modifying both the signature and specific predicates and functions. Results. A method for synthesizing conceptual logical Petri nets are implemented and illustrated with examples based on identifying the general semantics of conceptual graphs and Petri nets, resulting in the construction of models with declarative, imperative and dynamic properties.

Purpose of research. The main goal of this work is to increase the efficiency of automated game learning based on cognitive modeling. Based on the methodology of system analysis and cognitive modeling of weakly structured situations, the structure of a gaming automated training complex is proposed, which can be used in training personnel in various subject areas. In our case, staff training was simulated in stereotypical and non-stereotypical situations.Methods. This work is based on the general provisions of systems theory and system analysis, mathematical graph theory (which is based on cognitive modeling). The main tool of cognitive modeling was the construction of fuzzy cognitive maps of Silov. A modification of the algorithm for calculating the main system indicators of a fuzzy cognitive map was proposed. Game modeling, based on business games, was used for automated learning. The concept of an operational game was introduced, then modeling of the development of some unfavorable situation was carried out using fuzzy cognitive maps..Results. he main result of this work is the method of cognitive modeling of information support for game-based automated learning. Based on the developed methodology, a game simulation simulation of the operational game "House Fire" was carried out, which was based on the construction of a fuzzy cognitive map, for which the main quantitative system indicators of mutual influence, consonance and dissonance were calculated.Conclusion: the developed methodology allows for game modeling of unfavorable (including emergency) situations, which in the future will ensure adequate behavior of students in real situations.

Purpose of research. Creation of a methodology for designing a prototype of a metacomputer distributed computing system, taking into account the current stage of the evolution of hardware and cloud-network software to provide users with the means to create applications with inter-program parallelism and the ability of components to work together.Methods. Logical models of artificial intelligence, semantic networks and conceptual graphs, agent-based technology, virtualization of network resources. The method of conducting a full-scale experiment was that when the application was launched for execution in a virtual agent-based metacomputer, a network infrastructure was used with remote access to the Fast Ethernet laboratory network via the Internet, and then time characteristics were measured.Results. A technique for designing cloud-network metacomputer systems and applications is proposed, and prototype middleware software based on multi-agent technology is created. The goal of the study has been achieved, since the developed agent-based environment allows the implementation of universal programming control structures - transition by one or more conditions, cycle, sequence, parallelization, for which executable conceptual specifications have been introduced.Conclusion. An approach to the implementation of a distributed metacomputer application in a computer network environment based on conceptual graphs describing the exchange of messages and data processing by software agents is proposed. The performance of the application under conditions of extraneous load on the network was demonstrated.

Background. It is proposed to develop agent-based network metacomputer systems and applications based on logical methods and related conceptual graphical models, which allows combining imperative and declarative methods when designing the functional architecture and software of a metacomputer. Formalized specifications for creating agent-based network applications based on conceptual and logical models of artificial intelligence are proposed. The term “metacomputer” is chosen to denote the network environment in which the action script is deployed. Another name is a cloudnetwork application, in principle it means the same thing, but it differs in the additional consideration of terminology from the field of modern network technologies in an explicit form. In connection with the growing importance of global computer networks in science and education, the problem of creating large-scale applications is relevant. A functional organization of metacomputer agent-based network distributed computing is proposed, which implements the main structures of distributed programming, where the network is actually considered as a computer with distributed program control based on the messagedriven computing paradigm, and not as a means of implementing the simple client-server or master-slave applications. The aim of the work is to increase the level of parallelism in data processing in metacomputer systems by organizing the pipeline movement of messages over the network. Materials and methods. Conceptual models, logical-algebraic operating models, logical Petri nets are used as the main methods. Results. Conceptual graphs of distributed algorithms and logical-algebraic operational expressions suitable for use as directly executable specifications are proposed, a method is developed for moving from conceptual graphs to executable specifications that define the functional architecture of a metacomputer. Simulation models for distributed algorithms have been developed. Conclusions. The practical implementation of the above concepts and models will increase the level of parallelism in the operation of agent-based virtual metacomputer systems due to the pipeline organization of message passing.

Background. Despite the fact that by now new scalable architectures have been developed for distributed computing systems containing hundreds and thousands of computers, the task of creating transparent software of the “middleware” level remains urgent. The technology retrospective chain includes cloud, grid and metacomputer technologies, as well as various utility computing options. The creation of Internet-scale metacomputer technologies, tested at the completion of a number of well-known research projects, is of particular interest. The organization of effective management of the huge amount of resources that are available in a network environment deployed over a large area is a large and difficult problem in the implementation of distributed computing. The object of the study is metacomputer systems implemented on the basis of global networks, and the subject of the study is the organization of control of global computational processes in networks based on metacomputer technology. The purpose of the study is an improvement of this technology, which will allow considering a network with a given communication infrastructure as a single resource with the possible organization of arbitrary distributed algorithms. Materials and methods. The studies are carried out on the basis of the construction and software implementation of a conceptual model of distributed computing, implemented in a virtual metacomputer environment, which is the result of the integration of network, grid and cloud systems with agent-based systems. Results. The organization of metacomputer- based agent-based network distributed computing, which implements the basic constructions of distributed programming, where the network is really considered as a computer with distributed program control (message-driven computing), but not as a means of implementing the simplest client-server or master-slave applications. Modules, or agents, of a distributed application are capable of operating both in a reactive mode, waiting for data to be received and control transfer, and in a proactive mode, requesting data and control from previous modules (agents). Conclusions. Experiments carried out on a real network made it possible to confirm the operability of a metacomputer application and its ability to scale and expand its functional capabilities up to the properties of cloud-based network technologies AaaS (Agent as a Service) and FaaS (Function as a Service).

The article considers a dialog service which performs a bot assistant functions for recording metrological measurements results. The project conducted research of one of the modern approach to developing chatbot — rule-based chatbot. The main approach’s features were considered and compared with analogous bot assistant in that article. The aim of the research is development of the dialog service model to ease and accelerate metrological measurements. As a result, the article proposed recommendations for developing a dialog system which performs a bot assistant with functions of performing and recording metrological measurements results typed by the user to the chat or using a voice assistant.

Background. At the current level of organization of distributed computing systems (DCS), it is necessary to take into account the mobility of components - computing nodes located on mobile platforms and software modules in the form of mobile agents. The organizing of the interactions of components in virtualized cloudnetwork DCS (CN DCS), the software of which is based on the platforms of mobile and stationary agents, and the hardware includes stationary and mobile computing nodes, is an urgent and complex problem. The object of the research is the functional architecture of the CN DCS. The subject of the research is a methodology for constructing a conceptual model of distributed computing in CN DCS. The aim of the study is to create a simple conceptual model (CM) of CN DCS, combining the properties of cloud and grid systems with the properties of multi-agent systems and suitable for the subsequent creation of software for applied and middleware levels of CN DCS by sequential detailing of the CM. Materials and methods. During the research process, a model of a network of multi-tape Turing machines has being built and formalized specifications of network nodes are developed based on the theory of networks of abstract modules and executable logical-algebraic models. Results. A generalized conceptual model of the functional architecture of an agent-oriented cloud-network DCS with variable structure and mobile software based on a network of multi-tape Turing machines is proposed. The new unified descriptions of the elements of the conceptual model - deterministic, non-deterministic and probabilistic Turing machines with variable configuration based on the apparatus of formalized specifications - networks of abstract modules are proposed. Conclusions. A method for constructing a generalized conceptual model of the functional architecture of an agent-based CN DCS with a variable structure and mobile software based on a network of Turing machines, which allows a developer to evaluate the properties and determine the composition of DCS software of this type is proposed. It is assumed that, in practice, such a model is also suitable for the implementation on its basis of prototype software for systems of distributed and parallel symbolic multiprocessing of data.

Many users need the organization of complex virtual topologies, not just a simple hierarchical or client-server architecture, which makes it difficult to solve problems associated with data risk management. The subject of research is the processes of managing message flows in distributed computing and data exchange in cloud-grid systems with variable system and functional architecture. The purpose of the research is to improve data risk management as a result of the organization of distributed computing and data exchange in the cloud-grid and mobile distributed computing systems with a variable structure. In order to solve these problems the conceptual model of the cloud-grid system with several third party auditors, group managers, cloud services provider, user groups and user databases was built. In contradistinction to another concepts, it is proposed to implement network computing as a service organized on the user demand and implemented in a computer network, where one or another virtual system architecture is created.

Background. The object of the research is the functional architecture of distributed computing systems with a variable (reconfigurable) structure characteristic of hybrid systems of cloud-network (grid) type. Despite the fact that Petri nets have long been studied both theoretically and practically, the methods of their interpretation continue to develop intensively. At present, the problem of embedding Petri nets in the architecture of distributed network applications used to implement global computing in modern mixed cloud, grid and cluster systems has not been sufficiently studied. It is shown that in modern studies, Petri nets are used mainly in the simulation of discrete systems and processes, and not as the basis for formalized specifications in the development of distributed applications. In this regard, the interpretation of Petri nets in applications to the functional architecture of distributed computing systems with a variable structure based on the network software of the intermediate class (middleware class) is relevant. The aim of the work was the integration of graphical representations of conceptual graphs, semantic networks, scenarios and Petri nets, which made it possible to create effective tools with graphical support for designing a functional architecture of distributed computing systems with a variable structure and, in particular, a cloudy architecture of the NCaaSoD type - Network Computing as a Service on Demand (network computing as a cloud service at the request of the user).Materials and methods. The conceptual models of distributed processes that are a graphical interpretation of the first-order predicate calculus are used. Conceptual graphs for distributed Petri nets of mixed type have been proposed, which allow describing computation processes in global computational networks with a view to their subsequent implementation. Results. Based on the integration of graphical representations of conceptual graphs, semantic networks, scenarios and Petri nets, conceptual representations of distributed reconfigurable Petri nets are proposed, allowing them to be directly integrated into the architecture of the computer network.Results. New conceptual-behavioral models based on conceptual graphs of distributed Petri nets have been proposed to determine the system and functional architectures of distributed computing systems with a variable structure provided to the user as a hybrid cloud-based network service; these models are distinguished by the possibility of operational reconfiguration and immediate execution.Conclusion. A method was proposed and formalized for embedding conceptual Petri nets into the architecture of cloud-networked computer systems such as NCaaSoD — network (cloud) computing as a service organized at the user's request. The rules for obtaining relations of connectivity between the positions and transitions of the Petri net, placed on the nodes of the physical computer network, are proposed.

The concept "network is a computer" that has been further developed in the form of a paradigm of cloud applications that have the properties of "multi-lease"and "live" database migration is considered. Of particular interest in this regard is the development of a middleware for large reconfigurable clustered server systems as part of the support of the "big data" concept. Therefore, an approach is being developed to design a reconfigurable and parametrically tuned system and functional architecture of distributed computing systems. In some cases, the implementation of this approach can provide increased efficiency and cost reduction of large software and hardware systems. The article categorizes distributed computing systems when taking into account the specifics of cloud, grid, cluster and other types of communal, parallel and distributed computations. It is recommended to use as a basis hybrid architectures that combine the positive properties of cloud, grid and cluster distributed computing systems. The concept of organization of distributed network computing as services, implemented at the client's request, is developed. A network model that describes the work of a hybrid cloud grid system with third-party auditors and cloud service providers, is proposed. A distinctive feature of the proposed model is that when implementing as a concept NCaaSoD (network architecture as a service on the user's request) before the execution of a request, such as Upload (uploading data to the cloud), meta-information about available resources in order to further create a virtual cluster from the available nodes of the network of the cloud service provider, and before requesting the download type request (request information from the cloud), information is requested about the nodes that store the results. Meta-information can be obtained not only from a third-party auditor, but also from a remote monitoring service. To further develop the proposed concepts to real technology, it is advisable to develop a methodology for transforming conceptual models of the system and functional architecture into specifications suitable for the formation of virtual topologies of a network computer such as NCaaSoD in a hybrid cloud environment.

The article considers some “directly executable” formal models that are suitable for the specification of computing and networking in the cloud environment and other networks which are similar to wireless networks MANET. These models can be easily programmed and implemented on computer networks.