Intelligence system using the concepts of representations for solving problems of goal-directed behavior

The aim of the research is the intellectual systems, focused on the use of cognitive mechanisms relating to the issues of formation and formalization of reality representation for solving the problems of goaloriented behavior. The interest to the use of cognitive mechanisms arose after it became evident that while using various achievements in the field of information technologies, the computer is inferior to a man in solving some intellectual problems. The article examines the possibility of using cognitive approach. This approach allows to reduce the volumes of the information being processed while working out the managerial solutions for unplanned, unexpected situations.

For the use of cognitive mechanisms the author considers the methods that ensure their application in the formation and structuring of concept-representations. When solving this task we use modern methods and technologies of а modular approach. So service-oriented architecture represents a modular approach to software development. This approach uses replaceable components that have standardized interfaces and communicate over standardized protocols. An agent-oriented approach is equally important for smart system developers. Compared to objects, agents are not entities that require specific methods from each other, but entities that request the operation of required actions. The agent is at a significantly higher level of complexity with respect to traditional objects in an object-oriented approach. Using these approaches and methods, the author considers the issues of building modules of intelligent systems with the integration of the computer paradigm and cognitive mechanisms.

The article presents the structure of a cybernetic system for solving relatively simple problems of goal-oriented behavior. The system uses a classical approach to the representation of reality – sensual images. These images are formed of features, the values of which are obtained from various sources (sensors, software applications) OTS. The modules which are necessary for forming conceptsrepresentations of reality from sensual images are shown. A concept representation is a generalized sense-visual image of an object or phenomenon. Concepts-representations are static and represent a reflection of a set of the brightest external, sensually perceived signs of a separate subject or a reality phenomenon. A demo example of the formation of concept representation is presented. A fragment of the knowledge base containing the generated concept representation is shown.

The application of the considered approach will allow us to approach not only the problem solution of the formation of conceptrepresentations, but also their use for the solution of the task of a goal-oriented behavior. In this task it is possible to allocate two stages. At the first stage control actions (commands) can be used, which can be planned and implemented without changes. At the second stage it will be possible to form generalized representations of commands. In this case, you will need to translate them into a specific, executable form. These mechanisms can be used in cybernetic systems to reduce the amount of processed information in decision-making, to overcome unplanned, unexpected situations. 

1. Anfilatov V.S., Emel’yanov A.A. (ed.), Kukushkin A.A. Sistemnyy analiz v upravlenii: Ucheb. posobie. Moscow: Finansy i statistika, 2002. (In Russ.)

2. Kuznetsov O.P. Kognitivnaya semantika i iskusstvennyy intellekt. Iskusstvennyy intellekt i prinyatie resheniy. 2012. No.4. P. 32-42. (In Russ.)

3. Trembach V.M., Kognitivnyy podkhod k sozdaniyu intellektual’nykh moduley organizatsionno-tekhnicheskikh sistem. Otkrytoe obrazovanie. 2017. No. 2. P. 78–87 (In Russ.)

4. Golenkov V.V., Gulyakina N.A. Printsipy postroeniya massovoy semanticheskoy tekhnologii komponentnogo proektirovaniya intellektual’nykh sistem. Otkrytye semanticheskie tekhnologii proektirovaniya intellektual’nykh sistem = Open Semantic Technologies for Intelligent Systems (OSTIS2011): materialy I Mezhdunar. nauchn. tekhn. konf. Minsk: BGUIR. 2011. P. 21–58. (In Russ.)

5. Dirk Krafzig, Karl Banke, and Dirk Slama. Enterprise SOA. Prentice Hall, 2005.

6. Tarasov V.B. Ot mnogoagentnykh sistem k intellektual’nym organizatsiyam: filosofiya, psikhologiya, informatika. Moscow: Editorial URSS. 2002. 352 p. (In Russ.)

7. Tel’nov Yu.F. Model’ mnogoagentnoy sistemy realizatsii informatsionno-obrazovatel’nogo prostranstva. Chetyrnadtsataya natsional’naya konferentsiya po iskusstvennomu intellektu s mezhdunarodnym uchastiem KII-2014 (24-27 September 2014, Kazan’, Russia): Trudy konferentsii. Vol.1. Kazan’: Izd-vo RITs «Shkola», 2014. P. 334–343. (In Russ.)

8. Trembach V.M. Mnogoagentnaya sistema dlya resheniya zada tselenapravlennogo povedeniya.. Chetyrnadtsataya natsional’naya konferentsiya po iskusstvennomu intellektu s mezhdunarodnym uchastiem KII 2014 (24-27 September 2014, Kazan’, Russia): Trudy konferentsii. Vol. 1. Kazan’: Izd-vo RITs “Shkola”, 2014. P. 344-353. (In Russ.)

9. Tel’nov Yu.F., Danilov A.V., Kazakov V.A. Programmnaya realizatsiya informatsionno- obrazovatel’nogo prostranstva na osnove mnogoagentnoy tekhnologii i ontologicheskogo podkhoda. Otkrytoe obrazovanie. 2015. No. 6. P. 73-82. (In Russ.)

10. Kuznetsov O.P. O kontseptual’noy semantike. Iskusstvennyy intellekt i prinyatie resheniy. 2014. No. 3. P. 32–42. (In Russ.)

11. Rosch E. Cognitive representations of semantic categories. Journal of Experimental Psychology, 1975. 104. P. 192–233.

12. Lakoff J. Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago. University of Chicago Press, 1987.

13. Lapaeva L.G., Bychenkov O.A., Rogatkin D.A. Neyrobiologiya, ponyatiynye kategorii yazyka i elementarnaya model’ mira robota. Pyatnadtsataya natsional’naya konferentsiya po iskusstvennomu intellektu s mezhdunarodnym uchastiem KII 2016 (3–7 October 2016, Smolensk, Russia): Trudy konferentsii. Vol. 2. Smolensk: Universum, 2016. P. 292–300. (In Russ.)

14. Chudova N.V. Kontseptual’noe opisanie kartiny mira v zadachakh modelirovaniya povedeniya. Iskusstvennyy intellekt i prinyatie resheniy. 2012. No. 2. (In Russ.)

15. Rogatkin D.A., Kulikov D.A., Ivlieva A.L. Tri vzglyada na sovremennye dannye neyronauk v interesakh intellektual’noy robototekhniki. Modeling of Artificial Intelligence, 2015. Vol. 6. Iss. 2. (In Russ.)

16. Chelpanov V.G. Uchebnik logiki Moscow: Nauchnaya Biblioteka, 2010. 128 p. (In Russ.)

17. Trembach V.M. Reshenie zadach upravleniya v organizatsionno-tekhnicheskikh sistemakh s ispol’zovaniem evolyutsioniruyushchikh znaniy: monografiya. Moscow: MESI, 2010. 236 p. (In Russ.)

18. Satton R.S. Barto E.G. Obuchenie s podkrepleniem. Tr. from Eng. Moscow: BINOM. Laboratoriya znaniy, 2011. 399 p. (In Russ.)

19. Gavrilova T. A., Kudryavtsev D. V., Muromtsev D. I. Inzheneriya znaniy. Modeli i metody: Uchebnik. Saint Petersburg: Izdatel’stvo «Lan’», 2016. 324 P. (In Russ.)

20. Rybina G.V. Osnovy postroeniya intellektual’nykh sistem: ucheb. posobie Moscow: Finansy i statistika. 2010. 432 p. (In Russ.)

21. Rybina G.V., Parondzhanov p.S. Tekhnologiya postroeniya dinamicheskikh intellektual’nykh sistem: Uchebnoe posobie. Moscow: NIYaU MIFI, 2011. 240 p. (In Russ.)

22. Solodov A.A., Solodova E.A Analiz dinamicheskikh kharakteristik sluchaynykh vozdeystviy v kognitivnykh sistemakh. Otkrytoe obrazovanie. No. 1. 2017. P. 4–13 (In Russ.)