Three applied projects were carried out aimed at creating the digital infrastructure of Donetsk region, the infrastructure of academic and educational institutions of the city Donetsk and, especially, Donetsk National University
I participated in the project “Integration of university information resources into the Ukrainian network of Institutions of Science, Education and Culture (URAN)” as co-director of the regional project. As a result of the project in 1992, Donetsk National University and a number of other educational and scientific institutions in the region participating in the project were linked into a common information network, received access to the resources of the regional network*, as well as access to the national (Ukraine) academic network.
*Babosyuk A.L., Bashkov E.A., Kargin A.A., Petrov L.P., Yurkina T.V. Access to information resources or the development of an intellectual portal of Donbass / Zv’yazok. - 2002. - No.1. P. 45-46.
The next project, funded by the university and with the support of the Donetsk regional branch of UkrTELECOM coordinated by A.L. Babosyuk and me, provided for the creation of a corporate information system of university**. As a result, in 1998 all campuses of Donetsk National University were linked by their own fiber-optic data transmission channels (more than 30 km of total length). Construction of LANs in all structural divisions of the university (departments, faculties, administrative units) was completed. A corporate network structure and administration system were created. All corporate users received high-speed Internet access.
**Kargin A.A., Khristianovsky V.V., Kozhemyakin Yu.A., Demin V.A., Organization and administration of information resources in the corporate network of DonNU / Information systems and technologies in transport of Ukraine - state, problems, prospects: Collection of articles of international science and technology conference - K.: NTU "KPI", 2003. - P. 125-132.
And another project related to the informatization of the university was “Automated information library system of Donetsk National University”. A catalog's database on the base of international standard UNIMARK and workstations for all departments of the university library, including the reader's workstation, were created. An online service for ordering literature and remote access to electronic library resources has been created. Automated information library system has been put into operation since 1997.
Under my academic leadership 7 research projects were completed, financed by the Ministry of Education and Science of Ukraine from the Fundamental Research Fund
In 1989-1995, as part of USSR State Scientific and Technical Program "Advanced Information Technologies" under my leadership a project "Creation of mathematical, information, linguistic models and software for the hierarchical rules-based system of decision-making and control, as typical groundwork for comprehensively automated flexible production workshop" was carried out. Project was funded by State Committee for Science and Technology of the USSR.
In 2001-2003, the systems of situational control were under research. The models of representation and processing knowledge for real-time AI application* was focused. An architecture has been proposed and a toolkit for automating the design of adaptive self-organizing situational control systems** has been created. The experience of creating event-directed situational control systems is generalized and a model of intelligent situational machines is proposed. The foundations of the theory of intelligent machines and the scientific direction of intelligent event-driven situational machines have been laid. Research Report 2003.
*Kargin A.A., Aslanov A.E. On the interpretation of sound sources in situational intelligent machines/Herald of the Donetsk national University. Series A. Natural sciences, Donetsk, DonNU, 2001, No. 1, pp. 297-307. (in russia)
**Kargin A.A., Petrenko T.G., Paramonov A.I. Hybrid model of knowledge representation and retrieval / Information and control systems in railway transport. KharDAZT, Kharkiv, No. 4, 5, 2004, pp. 103-106. (in russia)
In 2004-2006, project "Intelligent machines based on integrated knowledge" had been conducted. The goal of the work is to develop a method for designing intelligent machines based on integrated knowledge. The method is aimed at automating the functions of a human who operates with sensory-figurative and verbal knowledge. The principles for interpreting fragments of the situation have been developed and formalized. The theory of interpreting the situation represented by hybrid data such as text, audio, and video data, as well as data from different types of sensors have been developed*. A conceptual model of the core of the interpretation system is proposed in which a situation formed from heterogeneous spatio-temporal fragments is homogeneously presented in the form of their meaning. The model of representing the sense has been done in the form of semantic relationships on a limited set of the domain concepts. This conceptual model of knowledge representation is formalized on the basis of the apparatus of fuzzy sets and fuzzy relationships. At this stage of the research, special attention is paid to the interpretation of textual information. There was three-level structure. At the top level, fuzzy characteristics of the of concepts confidences are formed. Concepts of this level reflect the meaning of the whole situation at a high level of abstraction. The meaning is formed on the basis of a sequence (stream) of concepts at a middle level of abstraction. At this level, the problem of aggregating heterogeneous data from different sources (text, audio, video and other sensors) in generalized solitary concepts was solved. At the lower level of interpretation of textual information, terms from the text were highlighted. This task included all the necessary stages of parsing text into sentences, words, establishing links, etc. Research Report 2006.
*Kargin A.A., Paramonov A.I. Hybrid fuzzy model of conceptual information processing/Proceedings of the 5th international conference "Intellectual analysis of information" - Kyiv, 2005. - P. 234-241. (in russia)
*Kargin A.A., Grigoriev A.V. Fuzzy search method for an unknown face using a photograph/Science works of the Donetsk nation. Technical Univ. "Computer technology and automation" series. Issue 88: Donetsk. RVA DonNTU, 2005, pp. 113-219. (in russia)
*Kargin A.A., Paramonov A.I. Intellectual Search System Based on Model of Natural Language Understanding / Proc. of the Third IEEE Workshop on Intellectual Data Acquisition and Advanced Computing Systems: Technology and Applications. Sofia, Bulgaria September 5-7, 2005. -P.150-155. (in russia)
Intelligent machines based on integrated knowledge were further developed in the next project 2006-2009 ”Intelligent self-learning machines based on integrated knowledge”. A prototype of an intelligent system for categorizing and interpreting textual information “Text-Term-Concept”* was created, which was experimentally tested on various tasks of interpreting heterogeneous data. One of these was the task of monitoring the operating modes of flour milling units based on the noise created by the equipment**. Interpretation of an audio data stream in real time, its classification, and alarming about pre- and emergency situations have been researched. The most important scientific result of the interpretation of heterogeneous data was possibility the real-time logging of the history of operating modes of units in the verbal form by using the concepts and relationships between them. Another task of interpreting heterogeneous data was processing video images of noisy and randomly oriented in 2-d space the text on the different images***. For text image recognition, a method and algorithm have been developed based on fuzzy cognitive models of human visual perception****. The model of adaptation through self-learning, based on cognitive learning theories, is of independent importance. Based on an analysis of the capabilities of learning models obtained in cognitive psychology, a model of an intelligent machine with self-learning is formalized. In these years several researchers concerning interpretation heterogenous data were conducted by my PhD students******. Research Report 2009.
*Kargin A.A., Lomonos Ya.G. Intelligent system for categorizing and interpreting text information "Text-Term-Concept"/Collection of scientific papers of the VII international scientific conference "Intellectual analysis of information" - Kyiv, 2007. - P. 140-149. (in russia)
**Kargin A.A., Aslanov A.E., Borukhson V.E., Petrenko T.G. Reducing the representation of fuzzy processes using the example of the problem of interpreting audio data/Information and control systems in railway transport. KharDAZT, Kharkiv, No. 3(30), 2001, pp. 5-8. (in russia)
***Kargin A.A., Pyatikop E.E. Application of fuzzy models of cognitive knowledge of image perception in the task of localizing text lines/Artificial Intelligence –2009. – No. 2. – P.161-167. (in russia)
****Kargin A.A., Petrenko T.G. Formalization of models of cognitive psychology based on the technology of situational intelligent machines (Part 1)/Bulletin of the Donetsk National University. Series A. Natural sciences, Donetsk, DonNU, 2006, No. 2. – pp. 349-353. (in russia)
*****Kargin A.A., Krachkovsky N.V. A cognitive model of the formation of prototypes in the task of self-learning of the situational control system/Electrical and computer systems. – 2013. – No. 9(85). – pp. 71-77. (in russia)
******Kargin A.A., Shariy T.V. Development of a framework for intellectual analysis of auditory data/Collection of proceedings of the VIII international conference “Intellectual analysis of information IAI-2008”, Kyiv, May 14-17, 2008, p.558-564. (in russia)
In 2010-2012, "Model of the cognitive level of the hierarchical system of information processing in the brain: an approach to computational intelligence" was the next step of development of intelligent machines. The model of the basic information processing unit (situation' granule) which is a generalization of information processes at the molecular (enzyme), cellular (neuron), cognitive (function) levels is proposed. The control system as a structured set of situation' granules mapping the context, motive and dynamic of situation have been described formally using the fuzzy certainty factor. On this basis, an architecture for an information processing system in the form of a multicomponent structure and a model for a classifier of fuzzy situations are proposed*. A description of the organization of the software and the results of computer experiments on controlling a LEGO robot based on the proposed technology are given. Research Report 2011.
*Kargin A.A., Krachkovsky N.V. Control of a mobile robot based on heterogeneous information/Bulletin of KhNTU, Kherson, – 2011.-№2(41).-P.339-340. (in russia)
*Kargin A.A., Krachkovsky N.V. Models of learning of a system of motivated contextual situational control/Vestnik KhNTU, Kherson, -2012. -No.1(44). -P.257-260. (in russia)
In 2013-1015, models, method and technology of processing information on molecular cellular level of living being have explored. Conceptual, mathematical and computer models which based on borrowed principles and verbal models of the interplays on molecular-cellular level of living nature* have been further developed. The models were tested on the following applications: a context-chains situational control system with self-learning for a mobile robot**, an intelligent racing controller in which the law of motion of the game's car is dynamically adjusted in accordance with the level of interest of the gameplayer, assessing the influence of poorly formalized factors in the formation of business plans***, formation of interaction model parameters for human-machine systems Research Report 2015.
*Kargin A.A., Petrenko T.G. About a new method of computational intelligence that borrows a model of the interaction of molecules inside a cell/Information and control systems in railway transport. KharDAZT, Kharkiv, No. 4 (Appendix), 2013, pp. 19-24.(in russia)
**Kargin A.A., Krachkovsky N.V. Method of computer-aided instruction in situation control systems/EUROPEAN RESEARCHER – 2013. № 9-1. – P. 2167-2178. (in russia)
***Kargin A.A., Babosiuk N.A. An approach to the formalization of the influence of verbal strategies on the criteria of a business plan based on models of computational intelligence/Artificial Intelligence - 2013. - No. 3(61). - P.324-331. (in russia)
In 2016-1018, models and method of processing information borrowed from living nature have been continued to study. The problem of creating machines combined features both smart and intelligent machines* is considered. Attention is focused on the limited possibilities of generalizing sensors data for making decisions in situations that were not laid down during the design of the system. The approach of extract knowledge from sensory data in the form of abstract categories (concepts) of different levels of generalization** is proposed. The mechanism of presenting data from sensors with abstract knowledge, differentiated by the level of generalization, is formalized in the form of a model of fuzzy granular computing. The approach is embodied in the pattern design method of IoT applications including smart things. A model of the new Smart Rules Engine pattern has been developed. The properties of the proposed method were investigated in a number of practical applications based on the hardware and software platform of the IoT&SM research polygon*** of Ukraine State University of Railway Transport. It has been experimentally confirmed the possibility of its use in embedded artificial intelligence systems for the mobile monitoring of the complex situations in real time. Research Report 2018.
*Kargin A.A., Petrenko T.G. Controlling a smart machine based on a model of categorical representation of a situation: a granular computing approach/Problems of information technology. – 2017. – No. 1(021). – P.17-28. (in russia)
**Kargin A.A., Petrenko T.G. Abstraction and categorization in smart machines based on granular computing/Bulletin of NTU "Khpi". Series: Informatics and modeling. – Kharkiv: NTU "KhPI". – 2017. – No. 50(1271). - pp. 57-68. (in Ukrainian)
***Anatolii Kargin, Oleksandr Ivaniuk, Georgii Galych Artem Panchenko. Polygon for smart machine application/Conference Proceedings of 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies DESSERT’2018 Ukraine, Kyiv, May 24-27, 2018, P. 489-494.