INTELLIGENT SYSTEM OF KNOWLEDGE SEARCH:
SYNERGETIC PARADIGM

V. Koleshko, A. Gulay, V. Gulay

Abstract. Phenomena of self-organization and self-rise of knowledge in the intelligent environment are considered on the basis of a synergetic model. The synergism base in development of the intelligent technology of the creative activity is represented by the cognitive universality of knowledge. Information from various cognitive practices is accounted by means of establishment of a multi-side frame model. The phenomenon of the synergetic development of the intelligent search system is considered by using an example of interaction of individual knowledge clusters.

Introduction. Phenomena of self-organization and self-rise of knowledge in the intelligent environment of the scientific search differ from simple summing-up the information in various communication systems by availability of spontaneity moments. Here, the spontaneity should be understood as the externally caused structural and functional transformations of the intelligent search system having its own internal driving forces. Mechanisms of such a drive generally occur in the intelligent environment of the scientific search as a complex system due to interaction of its components.

The said phenomena are considered on the basis of a synergetic model of the knowledge formation process. The synergism base in development of the intelligent technology of the creative activity is represented by the cognitive universality of knowledge which makes it possible to form most profound ideas on the studied subject. Information from various cognitive practices is accounted by means of establishment of a multi-side frame model of the intelligent search for knowledge. The phenomenon of the synergetic development of the intelligent search system is considered by using an example of interaction of individual knowledge clusters from different cognitive practices.

Establishment of a summarized phenomenon image in a self-developing intelligent environment. Presently, the issue of consideration and account of various cognitive practices of obtaining the knowledge and various options of its formation and use is becoming most urgent in the theory of learning. Importance of processes of knowledge generalization and synthesis increases due to the fact that in the 20^th century not only many new cognitive practices appeared, but rather non-traditional philosophic comprehension of well-known cognitive practices was noted. In the modern philosophy methodological prerequisites of synthesis of various practices and the use the experience of special epistemologies: social, religious, moral, economical, etc., have been laid down. The experience of knowledge studying and learning activity with the use of computer based intelligent technologies is to specially be noted. Today, synthetic knowledge problems are occupying the central position for understanding the variety of the surrounding world, study and comprehension of the society and a human being.

Within the frameworks of the scientific search intelligent environment complex practices are analyzed, wherein knowledge functioning is considered in such activities as designing, planning, management, studying. The diversified knowledge typology leads to account of its more frequent variants in the practical-methodological, natural scientific, engineering-technological and humanitarian knowledge. For example, the engineering knowledge unites diagnostics, control, adjustment; new variants and branches of the engineering knowledge are introduced: for example, programming, simulation. Such approaches as interpretation, explanation, verification, forecast, reduction have been introduced to the scientific search.

The principle of knowledge separation to the following categories: superficial knowledge (the knowledge of visible interrelations between individual events and facts); profound knowledge (abstractions, analogies, equivalents reflecting the structure and nature of phenomena) has been put to the base of models development and methods of knowledge extraction and structuring. Only the profound knowledge sufficiently explains various effects and they may be used for forecasting the properties and behavior of the studied subjects. Therefore, a necessity arise of establishing the universal mechanisms (schemes, methods) making it possible to reveal the profound knowledge strata referred to the studied area.

One of the most productive technology of knowledge structural designation is based on the use of a network of frames as abstract images having a certain set of attributes, as well as formalized models of these images. A remarkable frame model property is its universal character allowing to imagine all multitude of knowledge of the world with the aid of frames-structures, roles, scenarios, situations used for simulation of phenomena, objects, properties, parameters, modes. When the names of other frames are used as non-filled meanings of attributes (slots) of certain frames, a networks of the frames is formed. Obtaining the meanings in the frames forming the networks takes place by means of non-vivid heritage of frame properties on a higher hierarchy level.

The union of various cognitive practices for getting the combined knowledge is possible by means of establishment of horizontal ties between different networks of frames (fig. 1) [1]. In the give case one may say about formation of a multi-side structure of frames, wherein individual networks of the frames are subsystems. Being so, obtaining the meanings by slots in the frames of one network takes place from the slots belonging to the frames in the networks of other cognitive practices. Importance of the horizontal ties in the frame model is revealed when the levels of information interaction of the system components are analyzed. Formation of the scientific search intelligent environment on the basis of the proposed model presupposes ability of information interaction of its components on the profound level.

The level of profound and universal comprehension and the analysis of reality directly depend on availability of the horizontal ties between the systems of frames of various cognitive practices. The general understanding of values by all components of the system is typical for the given intelligent system, and it possesses a general thesaurus on the profound semantics level. The observed events and facts, discovered properties and dependencies are comprehended by the given system (in the form of information and knowledge clusters) in the context of value categories. Every cluster is interpreted as an individual stage of implementation of the profound idea, and it is analyzed from the point of view of its value within the framework of the implemented program of the scientific search or from the point of view of achievement of the earmarked goal.

Fig. 1. Formation of a summarized information image of the studied phenomenon in the intelligent environment.

The knowledge interpretation function is effectively carried out due to the horizontal ties, including ones obtained from other cognitive practices. The matter is that during data interpretation ability is realized of understanding the profound semantics of the information circulated within the system, finding out the profound sense in received and recorded facts. The interpretation role is especially big at the stage of the value orientation of the scientific search, when the goal of the intelligent system activity has not been preset too rigid. On the basis of interpretation of knowledge of various cognitive practices new scientific concepts are laid up by making it possible to see the general profound sense in abundant accumulated facts and to stimulate the precise search of new data. Fulfilment of the aforesaid intelligent system functions of interpreting and explaining the data makes it possible to ensure getting most productive ideas in mutually dependent and mutually influencing scientific fields and cognitive practices.

Synergetic model of the knowledge search intelligent technology development process. In equally balanced systems individual clusters of knowledge are considered as not linked with each other, every cluster exists actually without any dependence on other knowledge fragments. The intelligent system transition to the unbalanced state sets up the system coherency: the knowledge obtained from various scientific fields and cognitive practices begin to interact, form rather close links. As the balance state disappears, coherency of knowledge mutual dependence greatly increases. Far from the balance state a cluster interacts with the knowledge system on the whole, and it respectively influences the search of the final solution of the problem under study. One may say that the final number of clusters (the limited knowledge volume) demonstrates coherent behaviour in spite of their random selection and optional internal organization of every cluster.

So, far reaching correlations are found in the intelligent system of knowledge search in the unbalanced state, and the system begins to act as a single whole. Diverse knowledge obtained from different cognitive practices stop to be independent, isolated, alienated from each other, a well agreed ensemble and a single system of the interrelated knowledge appear. The said processes result in the situation when suppositions and versions of different degrees of trueness are rebuilt, approached and absorbed by each other, and finally, a part of them is excluded from consideration and analysis. This results in considerable reduction of a number of freedom degrees in interpreting the facts and data, i.e. stabilization of the formed knowledge system in the intelligent environment.

The totality of bifurcation phenomena in synergism is a fundamental mechanism of non-linearity of development of scientific search intelligent systems. In our consideration the bifurcation points are, most probably, critical moments of the research, when the accumulated sufficient knowledge volume determines the search of new beginning points for the further development of the creative process. For example, a critical moment in the intelligent environment includes achievement of a certain intermediate and particular result, after what selection of a new search trend is required along with attraction of another set of information and methods of its procession. A bifurcation point is also considered to be a moment of determination of erroneous movement to a certain direction in the intelligent environment, when the return is required to one of the previous intermediate points of the scientific search.

During construction of an analytical model of the intelligent system synergism we are going to suppose that one subcluster (fragment) of knowledge taken from the field of science under consideration (a scientific trend, a cognitological practice) interacts with one subcluster of knowledge from another field for establishment (synthesis) of one additional cluster of knowledge (fig. 2, A) [2]. Suppose, the clusters taken from two different scientific trends correspond to р₁ and р₂ of active subclusters participating in establishment of new knowledge. He total number of clusters N_S during the union of two scientific trends will be equal to:

N_S = N₁ + N₂ + N₃, (1)

wherein N₁, N₂ are the number of clusters introduced to the process of new knowledge synthesis from both scientific trends (fig. 2, B). Here, N₃ is the minimal number of additional clusters due to interaction of active knowledge fragments:

N₃ = min {р₁N₁; р₂N₂}, (2)

wherein р₁N₁, р₂N₂ are the numbers of active subclusters obtained from two interdependent fields of science.

In the given case the synergism factor depicting the relation of the quantity of all clusters of knowledge introduced for consideration from both fields of science to the quantity of clusters expected during summing up the knowledge formed at every science direction regardless of each other can be shown in the following way:

k = N_S/(N₁ + N₂) =  1 + (min{р₁N₁; р₂N₂})/(N₁ + N₂). (3)

Fig. 2. Synergetic model of the creative search development in the intelligent system volume.

If one of the scientific fields (cognitive practices) providing its resources for development of new knowledge is far more active than another field of science in producing the fragments of knowledge participating in their synthesis, i.e. р₁N₁ << р₂N₂ or р₁N₁ >> р₂N₂, then we obtain from (3) respectively:

k₁ » 1 + р₁/(1 + N₂/N₁) or k₂ » 1 + р₂/(1 + N₁/N₂). (4)

So, the synergism level during interaction of knowledge of various fields (cognitive practices) depends on the number of active fragments (subclusters) introduced at various stages to the synthesis process.

References

1. Koleshko V. M., Gulay A. V., Gulay V. A., 2008. Vestnik BNTU, № 6, pp. 72–80.

2. Koleshko V. M., Gulay A. V., Gulay V. A., 2009. Theoretical and applied mechanics, BNTU. Minsk, issue 24, pp. 44–57.