На шляху до вирішення проблеми "Semantic Web – Data Base"
The present state of the problem of "Semantic Web – Data Base" is analyzed. "Semantic Web" is analyzed from the standpoint of the integration approach, covering the results of research in the fields of neurophysiology, psychology, philosophy; it allows to formally...
Збережено в:
| Дата: | 2019 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | Англійська |
| Опубліковано: |
The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
2019
|
| Теми: | |
| Онлайн доступ: | https://journal.iasa.kpi.ua/article/view/175555 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | System research and information technologies |
| Завантажити файл: |  |
Репозитарії
System research and information technologies| _version_ | 1866302558145871872 |
|---|---|
| author | Kislenko, Y. I. |
| author_facet | Kislenko, Y. I. |
| author_sort | Kislenko, Y. I. |
| baseUrl_str | http://journal.iasa.kpi.ua/oai |
| collection | OJS |
| datestamp_date | 2019-08-27T22:12:50Z |
| description | The present state of the problem of "Semantic Web – Data Base" is analyzed. "Semantic Web" is analyzed from the standpoint of the integration approach, covering the results of research in the fields of neurophysiology, psychology, philosophy; it allows to formally define the quantum of knowledge as a separate situation of the visual level and clearly define the scheme of its verbalization in the form of a basic semantic-syntactic structure. The main result of the integration approach is presented by the following thesis: “The structural level of the linguistic organization is derived from the structural and functional level of the neural organization of the visual path”. From here we have a productive conclusion: the structural and functional level of the language organization will be the same for all languages. The second component of the "Data Base" covers (should cover) the entire social cognitive potential of knowledge, presented by a plurality of accumulated texts. The structured level of knowledge base organization is presented by a very small fragment of the neural network, which reproduces a separate situation of text information, but which, through a plurality of separate tokens, of its constituents (with corresponding references to other structural formations) forms a practically cognitive neural network of a certain knowledge area. |
| doi_str_mv | 10.20535/SRIT.2308-8893.2019.2.10 |
| first_indexed | 2025-07-17T10:26:05Z |
| format | Article |
| fulltext |
Y.I. Kislenko, 2019
115 ISSN 1681–6048 System Research & Information Technologies, 2019, № 2
TIДC
НОВІ МЕТОДИ В СИСТЕМНОМУ АНАЛІЗІ,
ІНФОРМАТИЦІ ТА ТЕОРІЇ ПРИЙНЯТТЯ РІШЕНЬ
UDC 004.81
DOI: 10.20535/SRIT.2308-8893.2019.2.10
ON THE WAY TO SOLVING THE PROBLEM
OF “SEMANTIC WEB – DATA BASE”
Y.I. KISLENKO
Abstract. The present state of the problem of “Semantic Web – Data Base” is ana-
lyzed. “Semantic Web” is analyzed from the standpoint of the integration approach,
covering the results of research in the fields of neurophysiology, psychology, phi-
losophy; it allows to formally define the quantum of knowledge as a separate situa-
tion of the visual level and clearly define the scheme of its verbalization in the form
of a basic semantic-syntactic structure. The main result of the integration approach
is presented by the following thesis: “The structural level of the linguistic organiza-
tion is derived from the structural and functional level of the neural organization of
the visual path”. From here we have a productive conclusion: the structural and
functional level of the language organization will be the same for all languages. The
second component of the “Data Base” covers (should cover) the entire social cogni-
tive potential of knowledge, presented by a plurality of accumulated texts. The struc-
tured level of knowledge base organization is presented by a very small fragment of
the neural network, which reproduces a separate situation of text information, but
which, through a plurality of separate tokens, of its constituents (with corresponding
references to other structural formations) forms a practically cognitive neural net-
work of a certain knowledge area.
Keywords: semantic Web, Data Base, integration approach, quantum of knowl-
edge, the basic semantic-syntactic structure.
THE PROBLEM OF “SEMANTIC WEB – DATA BASE”
Tim Berners-Lee has been working on creation of the WWW web for over ten
years. The main idea is to use a variety of agents for carrying out multiple tasks of
users creating separate tracks between the data bits stored on different computers.
Practically, this is the distributed system, which enables access to the variety of
interrelated documents through the Internet. With the advent of WWW, which to
some extent updated the hopes on modeling intellectual capabilities of human, the
ideology, presented by the interrelation “Semantic Web - Data Base”, which has
been heating the imagination of many researchers and generations up until now,
has been created. The first part is related to the problems of perception and “un-
derstanding” of a message, while the second deals with generation and utilization
of basic knowledge.
In the general case, those are the components of human cognitive potential,
which embrace the whole communication process, considering all its constituents:
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 116
perception of a message, understanding its content, making a decision, verbaliza-
tion of the decision, etc., which in general are connected with the first part of the
interrelation – Semantic Web. For its time it was a revolutionary step, which dra-
matically influenced the formation of the whole cluster of information technolo-
gies, oriented on modeling human speech activity. However, it is worth mention-
ing that this aspiring program stimulates the search and development of certain
directions of modeling the speech activity, and at the same time, it is still pretty
far from cognitive abilities of alive neurosubstance.
The second component – “Data Base” as a part of human cognitive potential
represented by global text information files, is practically formed in computer
networks, while the first component – “Semantic Web”, responsible for “under-
standing” the message, and is still on the stage of initial search. Nevertheless, Tim
Berners-Lee virtually suggested that realization of this ambitious intellectual pro-
ject might happen in next twenty years. However, time is passing and the number
of unresolved questions does not decrease, it even seems to increase, especially
regarding semantics.
In the wake of the euphoria of solving these ambitious perspectives, entire
groups, associations, institutes have been formed; presentations and polls have
been conducted, etc. The results of surveys conducted by a group of researchers at
Elon University (Janna Quitney Anderson — Pew Research Center's) and the
American Life Project (May 4, 2010) are considered to be quite interesting. The
main result is quite presentable, though cautious. The results of the survey con-
ducted by the research centers on this project are quite revealing.
895 respondents were selected; they would have spoken about the possibility
of realizing this project within certain timeframes. Evaluation of the results are
the following (based on [1]):
about 47% of respondents expressed moderate expectations regarding the
possibility of realizing the project: “The project would not be as efficient as ex-
pected and the average users would not feel substantial difference when the pro-
ject is realized” – this is the point of professionals who understand the complexity
of the issue;
almost 41% expressed hope that “by 2020 Semantic Web would succeed
and would be able to provide better service for average user” — this is the point
of those users who are quite unaware of situation and satisfied with the current
condition of the Internet;
the rest (12 %) did not express anything regarding the evaluation of this
project.
Thus we can see that the society in general especially the professionals has
aken a rather reserved view of the project and further development is nothing
more than unfounded illusions regarding future perspectives. Of course, the prob-
lem posed and voiced by Tim Berners-Lee refers to one of the greatest challenges
– modeling human speech activity, which is mainly realized by our neurosub-
stance.
We see that the problem of finding information in the gigantic repositories of
data stored on electronic media is more or less solved, and partially satisfies the
users who have extracted information either through libraries or in dusty reposito-
ries, fluttering innumerable folios in the pre-computer era. Yet, in our opinion, the
problem of meaningful processing of information remains a priority for humanity.
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 117
Can we shift this creative function on to the shoulders of the computer? Naturally,
the question arises, where does the problem lie, in which direction to go, and in
general — are these issues solvable?
Practically, this program (to our mind) is intended for modeling the proc-
esses of using (understanding) of natural linguistic information, which in general
quite often is “illogical and somehow playful and mistaken” and oriented on cog-
nitive potential of interlocutor. The last remark is that current natural linguistic
technologies are restricted to using them as key words. The key point is that our
language is still not analyzed enough; there are still a lot of problem and uncertain
issues. Practically, this is the main argument regarding the possibility of formal-
ization of natural linguistic technologies. Obviously, various questions arise, such
as what is the reason, why our knowledge about language are so poorly formal-
ized, why, …, why?
The Semantic Web — Data Base program attracts more and more attention
to the problems of speech modeling, on the one hand, due to the wider use of
computer technology for applied linguistic tasks, and on the other hand, as it de-
fines a new vector for the development and use of information technologies in the
field of modeling information processes in our neurosubstance. That means, we
already face the problem of modeling quite complex psychological functions that
could not be implemented at previous stages of the development of information
technology. First of all, it concerns the modeling of speech activity.
INDIVIDUAL LANGUAGE SYSTEM
In the context of analyzing the interrelation “Semantic Web — Data Base” it is
worth addressing to the accomplishments of remarkable Kyiver Lev Shcherba,
who since his birth (since being six months old) has lived in Kyiv with his par-
ents, graduated from the lyceum, studied on the first course of Kyiv university.
Further, he proceeded studying in St. Petersburg, where under the supervision of
Jan Baudouin de Courtenay he was conquering the highest ranks of European lin-
guistic science. Shcherba’s collection of works “Language system and speech ac-
tivity” and the work “On a three-fold aspect of linguistic phenomena and on ex-
periments in linguistics” in particular are consider the highlight of his linguistic
experience and future foresight [2]).
The main scientific work of L. Shcherba is the concept of individual lan-
guage system (ILS) as a combination of linguistic processor (LP), responsible for
structural and functional level of message organization, and basic knowledge
(BK), where the complete cognitive potential of a subject presented on language
level is stored and accumulated. The key distinction of human “individual lan-
guage system” and computer models is that nowadays in information technologies
all the cognitive potential (basic knowledge) is still presented on language level as
fragmented variety of text information files, while in human system this potential
is integrated to a single medium, presented by our neurosubstance. Nevertheless,
we are fully aware of certain structural identity between Tim Berners-Lee’s
method and Shcherba’s concept of ILS, where linguistic processor, responsible
for structural and functional level of linguistic organization and provides the func-
tion of “understanding” of a message, and basic knowledge is responsible for the
whole human cognitive potential. Note that the ISL concept was introduced by L.
Shcherba back in1927, i.e. almost a hundred years before there appeared a chance
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 118
to shift our linguistic competence on to the shoulders of computers. Well, let’s use
this ILS ideology as a combination of LP and BK for solving current problems of
modeling human speech activity, create their corresponding models 21, MM , and
lock them out on each other with the help bilateral connections (fig. 1).
In general, the LP is responsible for identifying the structural and functional
level of the organization of a separate message, while the basic knowledge (BP) is
considered a repository of all cognitive potential, taking into account the struc-
tural and functional level of the linguistic organization, but this time it is the cog-
nitive potential of a particular individual. We emphasize once more that there is a
difference between human cognitive potentials and computer networks: at pre-
sent, cognitive potential is represented in natural linguistic information technolo-
gies (NLITs) only by significant volumes of text (or speech) information, while
human cognitive potential comprises also the information from all sensory levels
(sight, hearing, touch, taste, smell).
When, apparently, everything is so transparent, the question arises, why
aren’t there no technologies for processing natural language information up to this
time? The reason is that (on the one hand) the language is such a complex and
multifaceted object for exploring, and especially for research, that it integrates
practically all directions of analysis of this phenomenon, starting with biology,
psychology, neurophysiology, philosophy, cybernetics, and other interdisciplinary
areas, (and on the other hand), language is still not sufficiently structured for for-
mal analysis and modeling.
The first author’s attempt to integrate diversified knowledge on linguistic
organization was in 1998, when on the occasion of century of establishing Kyiv
Polytechnic Institute (KPI) the author’s study guides “Language Architecture” [3]
was published. In this guides, from the perspective of statistical average level of
analysis of current state of structural linguistic organization (quite reflexively),
the notion of “basic semantic and syntactic structure” “(BSSS)” as a basic struc-
tural element of linguistic organization was introduced. However, the research
was some kind of generalization of that time knowledge regarding structural level
of linguistic organization, and did not reveal anything new in the sphere of lin-
guistics.
Fig.1. Individual language system: LP — linguistic processor; BK — natural linguistic
basic knowledge
Linguistic processor
Syntax
Morphology
DB of service
info
Model
of the world
“I am”
Text
synthesis
analisis
E
C
A
F
R
E
T
N
I
M1
M2
Vocabularies
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 119
The qualitative leap of author's approach to structural and functional level of
linguistic organization is based on many years of teaching the course “Sensory
Systems” at the Department of Technical Cybernetics at Igor Sikorsky Kyiv Poly-
technic Institute (KPI), during which the structural and functional organization of
all systems of sensation, including the stages of perception and processing of arbi-
trary information, were carefully considered. The significant moment was the fa-
miliarization with the works of Semir Zeki, concerning the structural and func-
tional level of the neuro organization of the visual pathway [4] and the work of J.
Hawkins “On Intellect” [5], summarizing the functional identity of all sensory
systems. Therefore, thorough analysis of the functional load of sensory systems,
taking into account the unresolved problems of linguistics, allowed synthesizing a
rather harmonious model of the formation and development of speech activity.
It is important to emphasize the significant stage of the author’s plunging
into neurophysiology of sensory systems after meeting and fruitful contacts with
the academician O. Kryshtal, which stimulated publication of the article “Neuro-
physiological bases of linguistic organization” in the NASU Reports (upon the
recommendation of O. Kryshtal) [6] and small but important author’s monograph
“From Thought to Knowledge”, published by Ukrainian chronicles in Kyiv in
2008 [7]. In these works the way from perceiving the information of visual level
to translating it onto language level has been thoroughly researched. All this al-
lowed making a solid conclusion that the structure of linguistic organization is
greatly defined by the level of structural and functional neuroorganisation of vis-
ual pathway.
The title page presents a procedure for translating information of a figurative
level onto language level in the form of a sequence of individual steps. In general,
visual information enters the retina, which is filled with receptors of two types:
sticks (in the number of 130 · 106 located on the periphery of the retina) and cones
(in the number of 6 · 106, filling the central foveal area, and capable of thorough
identification of the color range). When we want to get a detailed look at some-
thing from the surrounding environment, we project this particle (with the help of
the lens muscles) to the central foveal area. Such fragment of the image will be
determined as a situation. The situation is a fragment of visual component of en-
vironment, which gets to the central “foveal area” of retina and is processed to its
full extent. In each particular situation all its components Obj/Subj, their dynam-
ics Mov, their attributes Attr (Obj/Subj), Attr (Mov) and the extent of those at-
tributes Attr (Attr).
In section 3.4 (Verbalization of visual information), taking into account the
results of neurophysiologists and psychologists research, the procedure for proc-
essing a particular situation of a figurative level with subsequent translation of its
results onto the language level is clearly traced. It is important to note that this is a
procedure for processing only one situation, and it is implemented with a fre-
quency of 25-75 Hz, tracking both the static and dynamic characteristics of the
components.
Only a part of visual component, which gets on the central foveal area, that
is a tenth part of a visual field, is thoroughly examined. However, it is processed
quite elaborately: through identifying all objects/subjects, their attributes and the
extent of these attributes, as well as identifying dynamic characteristics of all the
defined components, with further identification of their attributes and the extent
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 120
of the attributes. This is, virtually, the scheme of processing of only one fragment
of an image on the retina – a situation. Although, in order to explore the visual
field totally we need to examine other situations as well. But on the way of evolu-
tion of visual pathway this function is realized in another way through using “sa-
cad”, the system of sporadic scanning of the retina completely in order to find
particular components. As we can see this algorithm is quite economical and we
don’t have to scan the retina field each time, it is enough to scan only those zones,
where the important (for the subject) components are. Thus, we examined the
whole path of processing both of certain situation for identification of all its com-
ponents with defining all objects, subjects, their attributes, extents of attributes,
and the complete image, getting to the retina. If it is true, there is still one ques-
tion — how the process of verbalization of information from sensory level to the
language one is carried out.
The first attempt by the author to integrate the accumulated diversified
knowledge of the linguistic organization (more precisely, about the stages of its
formation and development) was presented in the form Pre-Conference (poster +
workshop talk) on summit at the BICA-13 conference in Grandotel, Kyiv, where
the integration platform of linguistic organization “Systematic approach to the
modeling of speech activity” was presented.
The comprehensive view of such approach was shortly in BICA-14 journal
by the article “Back to basics of speech activity” [8]. Practically, it meant the
presentation of author’s vision of the problem of linguistic organization to Euro-
pean market. It is important that an extensive axiomatics of speech activity, which
turned to be not only possible but also quite important for understanding diversi-
fied functioning of speech activity, was presented. The next work of Yu. Kys-
lenko and postgraduate D. Serheev [9] was devoted to the problems of modeling
of knowledge base on the mentioned principles.
INTEGRATION APPROACH TO THE ANALYSIS OF LINGUISTIC
ORGANISATION
Tim Berners-Lee suggested the Semantic Web – Data Base interrelation about
twenty years ago, but the sense of progress in its implementation is still not very
visible. Why is this so? On the one hand, this is due to the fact that the problem
itself touches upon the issues of modeling complex cognitive processes occurring
in our neural circuit, and on the other hand, probably, when the problem is clearly
defined and voiced, presumably there are already certain horizons (achievements
and certain hopes) to solve it. Let's try to evaluate the prospects of solving the
problem more substantially, taking into account those developments that have
already been implemented. In particular, the author's last work “Personalized cog-
nitive feedback as a powerful lever for accelerated social development" [10] is in
some way related to the analysis of the cognitive processes that occur in our neu-
ral circuit and determine the progressive accumulation of cognitive potential. It is
necessary to clearly identify the platform from which the study begins, in order to
clearly understand how and where to go in this direction. Since the problem of
“Semantic Web – Data Base” is related to modeling of cognitive processes in our
neural circuit, and for their analysis, we have only a speech level available, then
we face the problem of analyzing human cognitive activity.
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 121
Current state of classical linguistics
The stated research direction is a rather ambitious project, and the author abso-
lutely soberly assesses its complexity. However, here we can trace such a solu-
tion: we should only turn to the analysis of the current stage of development of
classical linguistics through the evaluation of modern achievements by recognized
representatives of this direction. We should only recall that the first qualified
study of the structural level of the linguistic organization was presented by “Gen-
eral rational grammar” (the grammar of Port-Royal in 1660) by the authorship of
the logician and philosopher Antoine Arno and the grammarian Claude Lanceleau
[11]. The grammar was formed according to the results of the analysis of the
structural level of the linguistic organization of significant volumes of texts,
mainly religious. We must pay tribute to the creators of this work, who “felt” and
identified the main features of linguistic organization in the form of a dichotomy
“simple / complex” sentences, and also identified a certain category of “word
combination” for which they did not have a clear definition at that time. Only
now, we can assert that such a structure occupies the status of a separate full-
fledged sentence, if we assume that the “missed” element of such a structure was
just mentioned in the previous sentence and is still activated for a certain time in
our neural circuit. This is practically a “saving” scheme of linguistic means, ori-
ented on listener.
The feedbacks from prominent linguists on the current state of the structural
level of the linguistic organization are presented in Table 1 from the work of
BICA-13. We see that practically the entire linguistic elite is rather restrained in
relation to the current state of our knowledge of the linguistic organization. Par-
ticular emphasis should be placed on the opinion of L. Astakhova (head of the
department of the German language of the Dnipropetrovsk State University), who
at one time clearly expressed her attitude to this problem: “Linguistic community
has long been ripe for rejecting the existing theories of a sentence; the object of
syntactic research is as well unknown” [12]. That is, these are key, painful issues
of classical linguistics, which are still waiting to be solved.
The revelation of B. Horodetskyi regarding the current state of linguistics
became an impetus for the formation of a coherent, holistic picture of the linguis-
tic organization. This is a person who has edited the collection “New in Foreign
Linguistics” for more than 10 years and kept the pulse on key pan- European lin-
guistic problems. His verdict regarding the prospect of a language organization
study looks very categorical, but rather constructive: “Many troubles in linguistics
are due to the fact that language is still considered a form of reflection of
“thought” rather than a scheme for the organization and presentation of knowl-
edge. So, from here we have a powerful conclusion: in the realm of classical lin-
guistics it is first necessary to deviate from the concept of "thought".
The revelation of B. Gorodetsky regarding the current state of linguistics be-
came an impulse for formation of coherent consistent picture of the linguistic or-
ganization. This is a person who has been editing the collection “New in Foreign
Linguistics” for more than 10 years and kept the pulse on key problems of Euro-
pean linguistics. His verdict on the perspectives of linguistic organization research
looks rather categorical, but still constructive: “A lot of linguistic problems are
related to the fact that a language is still considered as a form of conveying a
thought rather than a scheme of organization and transferring the knowledge [13].
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 122
Therefore, hence we have a powerful conclusion that in the realm of classical lin-
guistics, it is first necessary to deviate from the concept of “thought” and master
and use the produc tive platform of “knowledge”.
Thus, for us, the generalized assessment of the current state of classical lin-
guistics and the way out of this critical state, when the concept of “knowledge” is
adopted as the basis of the research, instead of the “thought”, suggested by B.
Gorodetskyi becomes the directional sign. This is a very appropriate suggestion,
since current linguistics still defines language as the scheme of reproduction of
“thought”, which is a standard scholastic situation, when one (not clearly defined
concept) is interpreted through another, which is also not so well defined. How-
ever, the question remains – what is defined as “knowledge”, how is it formed for
a human and how is it used?
T a b l e 1 . Current state of classical linguistics
Authors Unresolved problems of classical linguistics
Whitney W.,
F. de Saussure
Language is a system of linguistic units of different levels
with no logical connection between them.
Piotrowski R.
(St. Petersburg)
Linguistics is not a theoretical science built on experimental data,
it is a exploitative science based on brief samples [20]
Astakhova L.
Dnipropetrovsk
state university)
Linguistic community has long been ripe
for rejecting the existing theories of a sentence;
the object of syntactic research is as well unknown [12]
Grammar – 70 All current knowledge of classical linguistics cannot be considered
from the perspective of a whole system [19]
Grammar
Port- Royal The problem of word combination still has not been resolved [11]
Gorodetsky B.
A lot of linguistic problems are related to the fact that a language is
still considered as a form of conveying a thought rather than
a scheme of organization and transferring the knowledge [13]
The answer is formed naturally and fairly transparently if one traces the en-
tire chain of perception of information by the sensory system and the successive
stages of its processing. Knowledge is the end product of a very powerful intellec-
tual activity of a person, covering the stages of perception and processing of in-
formation by the sensory system, with subsequent identification of the obtained
results in the cerebral cortex with the subsequent possible translation it onto the
language level. This definition embraces the entire chain of transformations from
figurative to language level, but further, we will explore the entire sequence of
stages more thoroughly.
Therefore, it is important to observe what is happening on the way of per-
ception, formation and use of linguistic information, but this is an appeal to neu-
rophysiology that is rapidly developing in the modern world and provides answers
to many questions. Figure 2 presents a difference in the approaches to the process
of synthesis of linguistic material. Classical linguistics was formed on a set of
finished texts, not plunging into solving problems regarding the nature of forma-
tion and processing of a linguistic message. By this time, it adheres to the idea
that “we perceive a human through language”. On the other hand, when we stand
in the position that “text is the end product of a rather powerful intellectual activ-
ity of a person” which necessarily includes the stages of taking into account the
dependencies of the perception of the environment by the sensory system (ranging
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 123
from retinal functions that determine all components of a particular situation and
ending with the functional load of the visual cortex, which defines all their attrib-
utes), we can find out what knowledge is, how it is formed and used.
Integration approach to structural level of linguistic organization covers the
results of recent researches of linguistic organization in various joint spheres,
such as biology (Haeckels law); neurophysiology of visual pathway (Hubel, Wie-
sel, Semir Zeki, Jeff Hawkins); psychology of perceiving the environment (defin-
ing a situation); philosophy (the structure of a message in the triunity of time,
space and action); cybernetics (ontogenetic parallels law) etc.
Defining the category of “knowledge”
We can present a certain sequence of language message formation taking into ac-
count the following steps. Our sight, like other sensor systems (Semir Zeki, [4], J.
Hawkins [5]), works discretely at 25–75 Hz frequencies. So, for the “knowledge
quantum” that is perceived and processed by a human, it is worth taking a sepa-
rate discrete (separate frame) of visual perception of the environment when it is
assumed that human perceives a lion’s share (80-90 percent) of all the informa-
tion that comes in general through sensory system through the visual system.
Speech activity is one of the most significant functions of human society
and, probably, one of the least explored areas of society’s existence and function-
ing. However, to date much knowledge in many areas of research of various
spheres of its operation has been received and accumulated. For that reason, while
exploring this issue it is worth “integrating” together all the knowledge received
in different spheres.
Despite author’s “solid experience” in the domain of the linguistic organiza-
tion (generally about 40 years, and over fifty publications), he managed in a cer-
Speech activity as a research object
Fig. 2. Integration approach to creation of linguistic organization: Defining principle —
“we perceive the language through human”
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 124
tain way to form a rather clear scheme for the establishment, formation and use of
speech activity.
Practically, halfway comprehensive and uncontroversial picture regarding
linguistic organization has been formed over the last decade, when the author
gradually started overcoming the problems of speech activity taking into consid-
eration current researches in various related fields. Relation of integration ap-
proach and classical linguistics is well illustrated by the slides from author’s work
for the Conference BICA-13 in the form Pre-Conference. On the slides, the com-
parative characteristics of integration approach regarding the long-established
accomplishments of modern linguistics. Classical linguistics adheres to the idea
that “we perceive a human through the language”, while the integration approach
changes the vector of research for the opposite – “we perceive language through
human”; that means we are going to consider the way how human perceive the
environment, how process, generates, and translates it on language level, as well
as how human uses it.
Such approach practically settles a lot of problems of classical approach to
structural level of linguistic organization since it traces the whole path of perceiv-
ing and processing of certain knowledge quantum through the visual pathway.
Thus, we are considering a “knowledge quantum” of visual level as a particular
“frame”, which we further define as “situation”. Taking into account the re-
searches of neurophysiologists of visual pathway, we can define this notion as
follows: “situation is a fragment of visual component of environment which gets
to the central foveal area of retina and is processed to its full extent”.
On average, to the “central foveal area” of retina (according to psycholo-
gists) only ten percent of the information perceived by the retina gets. However,
when about a hundred objects / subjects fall into the retina, in general, we see that
no more than ten components gets to the central foveal area, which are then proc-
essed fully, and after that, they can be translated into the language level. Surpris-
ingly this number totally coincides with the statistical data of linguists, who state
that the length of a simple sentence does not exceed seven plus or minus two
components. The general scheme of processing information through the visual
pathway was well illustrated in author’s monograph “From thought to knowl-
edge” [7], in which the schematic overview of the sequence of stages on the way
to realizing the interrelation “Reality–Text” was presented (fig. 3).
Parallel
processing
Retina
image of the
environment
Environment
Series
parallel
processing
Current
visual
fields
Sequential
processing
Textual
description
of the
environment
R
et
in
a
Im
ag
e
pr
oc
es
so
r
L
in
gu
is
tic
p
ro
ce
ss
or
Fig. 3. The stages of processing the environment in the direction “Reality–Text”: D —
the environment getting to the retina; S — the variety of situations, that could be proc-
essed sequentially; D* — identifying situations; IP — image processing; D** — image
level of processing the situations; LP — linguistic processor; D*** — symbol level of
presenting the environment
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 125
It is important to mention the perception of the environment through visual
pathway, hence processing of received information, is conducted with a frequency
of 25–75Hz. Such scheme ensures clear identification of not only static, but also
dynamic objects.
Stages of formation of the “Basic semantic and syntactic structure”
How does the individual situation is processed by the visual pathway? — Careful
researches of neurophysiologists give the answer to this question. According to
Semir Zeki [4], who has been summarizing the results of studies of the visual
pathway during the last half of the century, we have:
experimentally-confirmed the presence of ensembles of neurons (hundreds
/ thousands), which select all objects / subjects (Obj / Subj) from a separate situa-
tion (the third level of retina;
experimentally-confirmed the presence of ensembles of neurons (hundreds
/ thousands), which, select dynamic components with movement identification
(Mov) from a separate situation (fourth, fifth levels);
experimentally-confirmed the presence of ensembles of neurons in the vis-
ual cortex of the brain (hundreds / thousands), which identify their attributes Attr
(Obj), Attr (Subj), Attr (Mov) for all found components (Obj / Subj) and their dy-
namical characteristics (Mov);
experimentally-confirmed the presence of ensembles of neurons in the cor-
tex of the brain (hundreds / thousands), which also determine the extent Attr
(Attr) for the found attributes.
So, neurophysiologists gave us a clear answer to how the process of treating
the visual pathway of a particular situation is performed: with defining all of its
components, identification of dynamic characteristics, as well as with the clear
ability to differentiate all the diversity of attributes. Therefore, after processing a
separate quantum of knowledge - situation, a person should only “mark” the com-
ponents of the situation determined by the visual pathway with corresponding
language labels. But this is a very difficult, hard and long process of language
formation in human society.
The answer to how is the situation translated into the linguistic level was
given by A.N. Gvozdev, who had comprehensively and very carefully been traced
the peculiarities of the process of forming the linguistic system of his son Eugene
for seven years. He did what often many parents try to do, but he was a person
with higher linguistic qualifications who systematically and purposefully com-
pared every new step of mastering the language with a specific environment. The
results of the research were published in the academic edition in 1949 with dedi-
cation to his son [14]. These observations show a clear (Fig. 4) staged scheme of
formation of the linguistic system of a child, which ends at about three years.
At this age, the child finishes the process of forming the main structural ele-
ment of linguistic organization — the basic semantic and syntactic structure
(BSSS), as the main scheme of verbalization of a particular situation. Researchers
of child's language even claim that at this period the child becomes a “professor
of linguistics” at the level of processing the household language level. But this is
a very difficult, important and responsible period both for parents and for the
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 126
child, when you need to calmly and effortlessly compare and correct each compo-
nent of every real situation (subject, object, action, time, space, reason, condition,
consequence etc.) with the language equivalents of their verbalization. We should
mention Masaru Ibuka and his bestseller “Kindergarten Is Too Late” [15], which
excites all parents and future mothers. Why is it “late”? – Because, by this time
child’s neural network is the most plastic and effectively “sucks” (like a sponge)
all information about the outside world. Over time, this ability becomes less ef-
fective, and experts argue that if a child has not mastered the language by five
years, then, in practice, he is not able to become a full member of the society. Fig.
4 represents this level in stages (1-5), when the child learns the procedure of ver-
balization of a particular situation; the last stage (stage 6) represents the process
of synthesizing the message at the polypredicative level.
The sequence of these stages, and considering their functional load, a clear
organization of the verbalization scheme of a particular situation emerges in the
form of a standard procedure which is mastered by a child at the age of 2,5-3
years. Basically (according to the definition) — the structure of the BSSS is a
two-component, monopredicative scheme describing any situation from the real
or virtual world, all components of which are updated at the attributive level.
Practically, the Fig. 4 presents the sequence of stages of the child’s learning the
message structure of any complexity.
The functional organization of the BSSS structure (monopredicative level) is
mastered quite easily and on time, while the polydynamic scheme of message or-
ganization is learned later and harder; quite confidently the child masters it at the
age of ten or twelve years. The peculiarity of formation of the linguistic order of
the polypredicative level is that the system of functional connections remains
adequate to the BSSS with certain transformations, and a whole structure of the
BSSS may be put in the place of each component and so on, proceeding to the
level of reproduction of the recursive scheme of the message organization. In
general, the way of formation of this structure is carefully analyzed by the work
“To the origins of speech activity” [8], and the structure of the BSSS itself is pre-
sented at Fig. 5.
Situation
O
ne
-
co
m
po
ne
nt
T
w
o-
co
m
po
ne
nt
Objective
level
of BSSS
Attributive
level
of BSSS
Situational level of
message according
to BSSS scheme
Subj
Mov
Time
Space
Figurative level Stages of formation of monopredicative BSSS level Polypredicative
level
R1 R1 R1 R1
R1
R1
r1 r1 r1
R1
Fig. 4. Staged scheme of formation of the linguistic system of a person
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 127
Forming messages at the polypredicative level
The BSSS structure (Subj – Pred) reflects only to the monopredicative level of
organization of a message, whereas the message can be formed generally on the
predicative level, which combines several BSSS structures. Generally, this will be
a structure similar to the BSSS structure (Fig. 3), where the whole structure of the
BSSS may be put in the place of a separate component. At the same time, only the
system of relations, which connects separate situations, changes. The issue of
forming messages at the polypredicative level is well researched by Yu. Kys-
lenko, A. Khimicha “The structural and functional level of organization of the
linguistic processor for informational natural and linguistic technologies” [16].
The fundamental difference between the formation of messages of the poly-
predicative level is the usage of another semantic load as components of a sepa-
rate structure, as well as systems of relations in comparison with the BSSS struc-
ture. When the situational relations of the BSSS structure determine actual
coordinates of the situation in the surrounding world or environment (time, space,
reason, condition, consequence, etc.) then, in case of the polypredicative type,
these coordinates are formed in relation to another situation, using a relative rela-
tion type: “where there is S1, then S2 occurs”, “when S1 is implemented, then S2
will be”, and so on. In practice, the polypredicative message has a strict organiza-
tion and is clearly identified by a linguistic processor. Transformation of predica-
tive and situational relations is clearly analyzed in the following publication.
Generally, the scheme of forming a message of the polypredicative level can re-
late not only to the system of relations of the BSSS structure, but also affect the
relation at the attributive level (let's recall a familiar example – “This is the house
that Jack built. This is the malt that lay in the house that Jack built...”). It is im-
portant that the structure of the polypredicative message is also clearly identified
by the linguistic processor.
It should be once again mentioned that if an environment is perceived by a
person as a sequence of individual situations, each of which is processed by the
visual system under the same scheme, then the structural organization of the lan-
Attr(Attr) Attr(Attr) Attr Attr Subj Predicator
R1
R2…Rn
R0
r1 R2 …rm
Fig. 5. Basic semantic and syntactic structure: Subj — subject; Pred — predicate; 0R —
main relations (mother-predicate); nRR 1 — predicative relation; nrr 1 — syntag-
matic relations; Predicator — core of the predicate
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 128
guage, almost regardless of nationality, color of the color of skin etc. should be
processed under a standard scheme. Of course, languages differ significantly in
terms of lexical composition, schemes of formation of polypredicative level, but
their structural and functional organization have many common issues.
The final part to the integration approach
When the integration approach to the linguistic organization determines that the
linguistic organization is derived from the structural and functional level of the
neural organization of the visual pathway of a person and does not depend on the
nationality, color of skin, place of residence, etc., but is determined only by the
human genome, then we come to a significate conclusion that at the neural level
each separate situation should be processed in the same way - with defining all
components of the environment OBJ / SUBJ, their movement Mov, with further
identification of all their attributes ATTR (Obj / Subj), ATTR (Mov) and also the
extend of the specified attributes ATTR (ATTR). That is, the study of a situation
at the structural and functional level of the neural organization of the visual path-
way occurs in the same way, regardless of nationality, place of residence, color of
skin, etc. The difference is in the linguistic organization of certain components
(language labels), which identify the results of processing of a situation, in the
levels of lexical identification of certain environmental components, in the order
of their use, and also at the level of formation of mono/poly- predicative struc-
tures. Actually, this level finishes the analysis of the structural level of a particular
situation, which is translated into the linguistic level by the BSSS structure.
Therefore, we have a significant conclusion that all the cognitive potential
received, generated and accumulated by humanity, at the linguistic level (regard-
less of the language), will be structured by almost the same way based on the se-
quence of the BSSS structures (see Fig. 4), despite certain differences in their or-
ganization at lexical level, and at the text formation level in general. So, we can
now talk about the organization and accumulation of a global cognitive knowl-
edge base with free access thereto. Certain developments in this field have already
been made.
PROSPECTS OF FORMATION OF SEMANTIC WEB
Information load of Semantic Web
The integration approach gave us a coherent system of the structural level of the
linguistic organization, which is based on the standard BSSS structure. This is the
syntax of the language organization of an arbitrary message. So, naturally, the
question arises - what is semantics, and how is it formed? In our opinion, seman-
tics (semantics of a message) is a system of binary relations between elements of
the BSSS structure, the order and direction of which are determined by the
scheme presented at Fig. 4.
If we take the BSSS structure as a basis and remove all the relations identi-
fied by the system of relations: 0R is the main relation “mother-predicate”, which
connects Subj and Predicat together; mRR ,,1 — relations that clearly identify
all the predicative components of the predicate; mrr ,,1 — situational relations
that determine the coordinates of a particular situation (time, space, reason, condi-
tions, consequence of actualization of a particular situation), what does remain?
Apparently, there is an unordered dump of lexeme with no semantic loading at the
entire message.
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 129
It is worth noting, however, that the semantic load of all components of a
situation is gradually realized, shaped and generalized by the child in the way of
mastering the linguistic organization. That is, the process of mastering the lan-
guage also means the formation of a semantic message system at a monopredica-
tive (age up to 3–5 years) and a polypredicative levels (age from about five to ten-
twelve years).
Still, the question remains of how does the recipient identify these attitudes
at the stage of perceiving the text? The answer will also likely be clear - either by
using a certain lexical load of components that determine, for example, time
(hours, minutes, seconds), stages of the day (morning, day, night, etc.), etc., or
through the link to other situations, which were uniquely identified earlier.
This small excursion to the semantic relations of the BSSS structure draws
the attention of the readers to the problem of formation and usage (“understand-
ing”) of semantic relations to search for relevant information; moreover, they are
clearly identified either in separate word forms or have a relative reference
scheme to other quantifies of knowledge. Well, we can proceed with analyzing
this problem.
So, there is the problem of synthesizing the Semantic Web in such a way
that, in part, it was possible to identify semantic links from textual information.
We perfectly (and almost automatically) identify these relations, because during
the long evolutionary way, we have worked out such schemes for identification of
semantic relations, which ensure adequate perception and reproduction of text at
the stages of analysis and synthesis of the message. Again, please remember that
the semantic load of a separate BSSS structure is determined by a system of func-
tional (semantic) relations of its constituents. So, we can predict the Semantic
Web for one component of the message presented by the BSSS structure. One should
only consider that for our language, the order of BSSS formation can be free.
We live in the period of computerization, we often quantify our thoughts
with “bits” or “bytes”, but not always use “knowledge”. The language, the writ-
ing, is a great achievement of humanity, it is something that distinguishes man
from animals. We have learned to record our thoughts, our experience for the
term of preservation, which far exceeds the life cycle of a person.
It is time to pass it to your computer, in other words - to put knowledge to
the machine, presented by a plurality of texts accumulated and saved by human-
ity. We can confidently identify the temporal and spatial relations, reason, condi-
tion, consequence, etc. in the texts. This knowledge should be “put” into the
memory of the computer in the form of a linguistic processor (LP) and used.
These are almost simple problems that can help programs “understand” the text
by identifying individual functional relations between the words. This is the main
meaning and purpose of the computer – not to lose the accumulated knowledge
and learn to understand and use it.
Again, we return to the problems of modeling the Individual Language Sys-
tem (ILS), which clearly separates two components: the linguistic processor (LP)
and the basic knowledge (BK). Let's carefully form the BK, where, at the textual
level, there will be formed all the major accumulated cognitive potential and
“gradually” we will form a linguistic LP processor to teach the “foolish” machine
to “understand” the text, gradually putting separate bricks of knowledge thereto,
presented by the plurality of interactions of the same BSSS structures. It is impor-
tant to capture the semantic load of the individual BSSS functional relations (to be
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 130
a reason, to be time, to be a place, etc.). The structure of such relations can be eas-
ily and unambiguously identified by a linguistic processor. All components of the
situation - Obj, Subj, predicate, situational relations (time, space, reason, conse-
quence, conditions) - are easily identifiable and work for us).
Let’s remember A.S. Narinjani and his school, which deals with problems of
“time”, “space”, etc., Let's also recall N. Leontiev, who argued that the texts al-
ways have all the information necessary for a thorough analysis of the message
and the use of this information. This issue is partly covered by to the publication
of the author “Structural and functional level of organization of the linguistic
processor”, which actually analyzes the means of identifying situational relations
for the mono/poly- predicative levels of linguistic communication organization
[17]. These works also presented a virtual experiment regarding the possibility of
searching the Internet for a “full text”. The main components here are components
of Subj-Pred, situational relations (time, space, movement), predicative relations.
It is important to emphasize that the system of relations of the BSSS struc-
ture at the mono/poly- predicative levels cements the set of components of the
“lexemes” into a monolithic knowledge quantum, which determines the “semantic
load” of a message. That is why special attention should be paid to connecting the
set of unsystematic lexemes to a semantically executed message. What do we
have now?
Comparative analysis of various schemes of formation of “Semantic Web”
The integration approach to the structural level of the linguistic organization,
presented in Section 3, gives a clear vision of the linguistic organization both at
the monoprostand level, which is confined exclusively to the BSSS structure, and
at the polypredicative level, covering a plurality of such structures, based on
general level of the recursive scheme of the organization of the message.
However, the basis of the structural level of the linguistic organization has always
been the basic semantic and syntactic structure — BSSS.
In practice, the basis of the entire structural level of the linguistic
organization is the basic structure, the semantic framework of which is de-
termined by the system of relations of the BSSS structure both situational and
predicative, presented at Fig. 4. Below is the list these relations:
ratio 0R — the main relation of the structure that identifies the connec-
tion “Subject-Predicate” — (Subj-Mov);
the predicate itself with the “Predicator” core, which covers the verb (or
its transformations according to the schemes of the adjectives or adverbs, based
on two groups of relations - situational and predicative;
situational relations nrrr ,,, 21 — determine time, space, reason, condi-
tions, consequances of action. . . in the environment;
the predictive relationships mRRR ,,, 21 R1, R2, ... Rm determine other
entities / objects that are involved in the actualization of the action identified by
the predictor Pred.
It is important to emphasize: the total number of elements that are
simultaneously defined by one situation does not exceed from seven to ten. This
is the average number of components that can simultaneously be processed by our
visual pathway. In case of a polypredicative message, this number will depend on
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 131
a plurality of sequentially reproduced situations. It is also worth considering the
number of attribute relations of all involved objects / subjects, when they are
identified as: Attr (Obj / Subj), Attr (Mov), Attr (Attr). This is a system of
components that uniquely identify the attributive level of a environmental
situation. The person is not able to perceive the situation in more details on a
figurative level and identify it, respectively, at the language one.
So, when we have determined the scheme of perception of a environmental
situation, and transform it into linguistic level, we can now correctly put the
question about the possibility and effectiveness of using this system of relations to
search for information in powerful natural and linguistics (NL) information
repositories.
Modern ways of presentation of Semantic Web
We have just studied how adequately, fully and thoroughly a person may perceive
environmental information, considering the semantic load of all components of a
situation and translate it into the language level. Here the process of the recipient
is actualized in the direction of “Reality – Text”, whereas for users (listeners /
readers) it is important to reproduce this dependence in the opposite direction –
“Text – Knowledge”.
The problem for the readers to apply to the gaining potential of knowledge
presented at the computer level, is solved so far, considering the artificial lan-
guages of NL texts. Such first attempt was implemented in the form of SGML
(Standart Generalized Markup Languuage), approved in the 80's when a certain
list of instructions (tags) was created to reproduce the text structure. HTML is a
simplified version of SGML structure that defined a certain set of Tags, their
attributes, and the internal structure of the text defined by the DTD rules for
certain types of document. However, it is important to emphasize that, in fact, the
texts themselves, as cognitive enhancements of our neural network, are by no
means connected with the Tags used for formal presentation of texts; nor does it
consider the semantic relations between Tags, the number of which is limited.
Considering the situation, the specialists have already expressed the idea that
HTML today does not fully meet the requirements of both the developers
themselves and the users. The proposed version of XML (according to the design
of the developers) itself should “synthesize” a specific set of tags by text; but this
means that it must have its own “intelligence” in all areas of knowledge.
Currently there is no answers of how to formalize this semantic problem!
Given these unresolved issues, the specialists state that: the evolution of data
structuring systems under the Obj-Attr scheme gradually brings researchers to the
complication of tags, which ultimately requires the ability to analyze the entire
structured level of text information.
Because of the analysis of the current state of the problems of searching
information on the Internet, considering the speed of computers, the possibility of
presentation of information in distributed systems there raises a question – should
one refer to a “full text” search? Certain proposals have already been stated [18],
although the author has not yet found this proposal.
The author is not an expert in computer technology but has achieved a good
level of generalization of the structural level of the linguistic organization, which
is derived from the structural and functional level of the neural organization of the
visual pathway, and therefore - will be the same for all languages at the structural
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 132
and functional level of their organization. In connection with this, there is a
suggestion: to transform the search system according to the BSSS standards, as a
generalized scheme of synthesis, analysis and presentation of a linguistic
organization regardless of a language.
Let's consider modern schemes of indexing the linguistic material, which are
used for searching in modern technologies.
Using triplestore
The indicated ratio covers a large amount of developments related to the
knowledge and modeling of two diverse ways of modeling the linguistic activity
of a person. The first component is Semantic Web, which is related to modeling
the process of “understanding” the text information, while the second — the Data
Base is associated with modeling the processes of accumulation and use of human
cognitive potential. So, it's worth differentiating these, interconnected problems,
to thoroughly analyze their features.
The concept Semantic Web was introduced by Sir Timothy Berners-Lee in
the early 2000's. Conceptually, this is the development of the Internet in the
direction of presentation of knowledge (again this uncertain term “knowledge”!).
By this time, Internet resources present this concept by unstructured WEB3.0 xb
WEB of DATA content, which is accompanied by certain metainformation.
A similar representation allows to create (intelligent / semantic) information sys-
tems that are oriented towards “understanding” the content. The ideology of
Semantic Web uses many formalisms fixed in standards.
The main scheme of information representation in Semantic Web is a triple,
which updates the combination of subject, predicate and object. The structure of
this representation is presented in Table 2.
T a b l e 2
Subj Predicate Object
my apartment has my computer
my apartment has my beg
my apartment is in Philadelphia
The concept of triple is so generalized that with the help of an unlimited
number of triples we can describe anything. Next, in Semantic Web, as well as in
general, in Computer Sciences it has been agreed to share data and data model
(metadata). In this case, the model / metadata are implemented in the form of the
ratio of ontology / taxonomy using OWL / RDFS, and the actual data are realized
is in the form of RDF-triples.
The peculiarity of Semantic Web is that there is no clearly defined structure,
as in the world of relational data, where the types of relations between the
elements are recorded and stored in accordance with the model. The structure is
quite complicated. In 2000, many different schemes of triplestores were created.
Triplestore is the foundation of Semantic Web.
MIVAR approach to forming the “Semantic Web – Data Base” model
Another approach to solving the problem of “Semantic Web - Data Base” was
proposed by O.O. Varlamov (Moscow), where for the presentation of text infor-
mation and search implementation, there was used another triade – “mivar” as a
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 133
ratio “If 1S , then 1S ”. This is, in practice, a model of metadata,
which, according to the author's idea, should solve the problems of the relation of
“Semantic Web – Data Base” in both directions:
on one hand, it is used to create, accumulate and use the global database
and the rules of their use based on adaptive discrete, mivar informative space
based on the triade “object, property, relationship”. It was assumed that the
technology could be used both for accumulation of knowledge (Data Base
formation) and for search of information (Semantic Web);
on the other hand, the mivar technology was developed to form a logical
conclusion using the ratio “If 1S then ... nS ...”, where nSS ,1 are related
situations connected by certain logical connections.
Somewhere at one of the conferences of the Seminar “Artificial Intelligence”
in Katsiveli (Crimea) the author managed to hear to the report O.O. Varlamov. At
that time, according to the speaker, there was accumulated a mivar base, which
included several million mivaras. A positive point is the use of logical conclusion
about the plurality if mivars. Such base can also be presented by the triples “If
1S , then ... nS ....”,
However, despite the principled opportunity to present the environment
through such triplets, the procedure itself turned out to be quite complicated and
unusual for users of the natural language. On the other hand, real text turned into
a meta-text that is not badly perceived by the computer, but people who were
engaged in this unnatural transformation of the text, had a lot of problems.
BSSS-option
It’s time to present the problem of searching the information under the “full text”,
which is based on the basic semantic and syntactic structure of the BSSS, as the
main and standard structure of the presentation of text information. So, it is worth
to compare the existing Semantic Web presentation schemes with the system of
relations of the BSSS structure (Fig. 5).
For the system “triples – store” we have a set of standard triples “Subj – Pred
– Obj”, through the sequence of which (according to users) it is possible to repro-
duce (present in a certain way) an arbitrary text.
For the mivar approach, on the one hand, the standard scheme of the triples
“Subj – Pred – Obj” is used, and for forming the logical conclusion, there is used
the sequence of triples of another kind “If 1S then ... nS ....” is used.
However, in both cases, experts who perform the procedure of indexing the
textual information under the unnatural scheme of text processing, express certain
claims to customers.
Let’s now look at the system of relations of the BSSS structure, which we
have mastered since childhood and perceive and use in a completely natural way.
The integration approach covers a system of simple binary relations:
Subj – Pred Subj – Attr(Subj) Attr(Subj) – Attr(Attr)
Pred – Obj1 Obj1 – Attr(Obj1) Attr(Obj1) – Attr(Attr)
Pred – Obj2 Obj2 – Attr(Obj2) Attr(Obj2) – Attr(Attr)
Pred – Objn Objn – Attr(Obj2) Attr(Objn) – Attr(Attr)
This is almost the system of relations of a separate structure of our native
language, which we use daily - it is part of our language: (noun, verb, adjective,
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 134
adverb), the combination of which form the standard BSSS structure of a separate
message. If to use the system of relations of the native language, which we per-
ceive automatically, then our head will not ache. However, for the linguistic proc-
essor to automatically identify them, one, probably, should work a bit.
PROSPECTS FOR THE FORMATION OF THE NATURAL LANGUAGE “DATA
BASE” ON THE PRINCIPLES OF NEUROPHYSIOLOGY
Then we will continue the study the process of forming the structural level of the
“Data Base”, considering certain features of the organization of our neural net-
work. The publication Kislenko–Sergeyev [9] presents a model of memory or-
ganization, where the cognitive element of the linguistic organization is the BSSS
structure itself. There arises, of course, the question – how can a set of cognitive
structures pass through a separate node of the neural network (analogue of the
neuron)?
Here we come to the concept of “lexeme”, which, in practice, presents a
model of a separate neuron, through which there can be many options for imple-
mentation of individual components of other BSSS. Lexema, practically defines
all possible variants of usage of a wordform, which ensures the possibility of its
entry to any level of the cognitive element of the linguistic message. For modern
computer equipment, this is a simple procedure when we have a whole bank of
lexemes for each language, as presented as the first step towards the formation of
a cognitive structure.
The next step – there is “built” a corresponding framework of the BSSS
structure (as an element of the cognitive web) based on the set of corresponding
lexemes, through which many cognitive connections may pass. Thus, a real text is
formed based on the separate of actant of lexemes, etc. (this is where the required
lexemes are not in the dictionary). Continuing this procedure, we form a separate
BSSS structure, through the “body” of which many connections may pass. That
is, a separate lexeme can be a constructive element in creating a set of other simi-
lar structures. This is practically the procedure for the formation of a separate
element of the BSSS structures, through which several other BSSS structures can
pass. In general, such model represents integrational properties of the biological
neuron.
To implement such functions, there was suggested to use the “marker” de-
vice (see Fig. 5); A separate BSSS structure, when reproducing a text, receives its
unique marker (identifier). Thus, we have a model for forming a cognitive net-
work of separate text, where each component of the message (except for BSSS)
has its own unique number. Well, the procedure is complicated – but it can be
performed automatically, freeing a person from this unskilled hard routine. This
practically means that, with a limited number of lexemes (vocabulary), we will be
able to reproduce virtually unlimited number of real texts. In a certain way, we
are trying to implement the model of information accumulation in our neural net-
work at the neural level. This is a virtually definite model of the neural network,
where on the plurality of models of individual “neurons – lexemes” can be actual-
ized a practically significant set of individual quants of knowledge in the form of
BSSS structures, reproducing the cognitive potential of individual fields of
knowledge – astronomy, mathematics, physics, etc.
Is it possible to implement such a cognitive neural network? - Basically –
Yes. When taking into account the possibilities of forming and using powerful
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 135
distributed systems for the following functions: preservation of dictionaries of
different languages, preservation of appropriate vocabularies of lexemes, forma-
tion of linguistic processors for the decomposition of arbitrary messages based on
the BSSS structures, automatic marker formation system for each BSSS structure,
a powerful translation system, where the elements of the translation are not words
– but separate situation ... and so on.
What do we have in the final version? There is a large list of components
for normal formation and operation of a whole cluster of Information Natural and
Linguistic Technologies (INLT); a clear list of important tasks, each of which is
being already solved as of date; there is only a lack in systematicity the entire
chain of formation, processing and use of natural and linguistic information is not
carefully researched.
Nevertheless, it is positive that these issues have already been put on the
agenda, certain fields of research and design have been carefully described and
investigated; young, ambitious. unbounded people join the development. It is also
important that the set of lexemes of a particular area of research may be synthe-
sized for a significant number of other quanta of knowledge, which provokes an
ambitious question regarding the prospects of the possibility of automatic synthe-
sis of new Knowledge. It is difficult, but this is probably the principle of the
global artificial neural network, which is developed by the achievements in many
scientific fields.
This model covers a plurality of models of individual quanta of knowledge
presented at the level of the set of BSSS structures, which removes many prob-
lems, since a single quantum of knowledge is equally presented in all languages.
At the same time, almost, all problems with implementation of such network are
removed, if the basis of its formation is determined by the real text without any
restrictions. Apparently, distributed information systems will be the basis for
solving the problem of “Semantic Web – Data Base”. We can present a fragment
of the neural organization of the proposed knowledge base structure (Fig. 6),
where a separate element of the BSSS structure is updated with an appropriate
lexeme associated with other lexemes of the reproduced text by a plurality of
unique markers.
The linguistic processor, which allocates separate BSSS structures (for mono
or poly- predicative organizations), translates them to the lexeme level, automati-
cally forming a unique marker for each situation. This is, basically, a model of
what is being implemented in our neural network on a plurality of neurons; this is
a model of a neuron through which many relations with other quanta of knowl-
edge can pass.
The main conclusion is considered rather optimistic: “there is proposed a
model of automatic formation of the cognitive potential of the community in the
form of a system of integration of all accumulated text material in various fields
of human activity (physics, mathematics, biology, etc.) based on the standard
BSSS structure as a unified scheme of textual information presentation”.
This means that we proceed to the constructive level of formation of the
“cognitive potential of the society” in various areas of human intellectual activity,
because we have structural capability of formation, accumulation and correction
of knowledge presented at the linguistic level by a plurality of reports, research
results, monographs, textbooks etc. Perhaps, the first research in this field is the
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 136
work of Toby Segaran “Programming the collective mind”, Symbol - Plus 368 p.
[18] (unfortunately, I have not yet been able to get acquainted with this work).
This is an interesting perspective direction of research, formation and generaliza-
tion of collective achievement of humanity in different fields of activity.
If the quantum of knowledge for the perception of a person of the environ-
ment presents a separate situation of a sensory level, and its verbalized form is the
BSSS structure, then it will be a “cognitive element” of perception, accumulation
and formation of the cognitive potential of the community, formed in separate
areas of scientific research.
In practice, this section is devoted to the problems of formation of the “Data
Base” on the principles of organization and operation of our neural network. In
many ways, the proposed model functionally resembles the work of our neurosub-
stantium: perceives, “understands” and accumulates linguistic (text) information.
SEARCH FOR THE SEMANTIC UNIT OF LINGUISTIC ACTIVITY
Basically, the previous section gave us an answer to this question. However, it is
important to analyze the way of solving this issue. The presumption of
formulation of such a question is determined by the rapid (powerful) development
of computer technologies and the use of their capabilities, even for the simulation
M1 M2 M3
M4 M5 … Mk
M1
M2 … Mp
M1
M2 … Mq
M1 M2
… Mr
M1
M2 … MS
M1 M2
… Mt M1
M2
… Mu
M4 M5 … Ml
M1 M2
M3
Attr(Attr) Attr(Attr) Attr AttrSubj Predicator
R1
R2 … Rn
R0
r1
r2 … rm
Fig. 6. The structure of the neural network fragment: Subj — BSSS subject; Attr (Subj)
— subject’s attribute; Attr(Attr) — extend of the attribute; Predicator — core of BSSS
structure; Attr(Pred) — predicative attribute; Attr(Attr(Pred) — extend of the attribute of
predicative
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 137
of the most complex intellectual functions of a person. However, before transfer-
ring individual functions of intellectual activity to a computer, it is necessary: to
clearly understand the statement of the problem and the question, the main princi-
ples of the linguistic organization, the scheme of language communication, repro-
duction of the environment by the means of language, “understand the message”,
etc., formalize these solutions for implementation at the computer level, or distri-
bution systems.
1. The first stage has almost been performed by Sir Timothy Berners-Lee –
he clearly set the problem of studying the linguistic activity as an acronym
“SEMANTIC WEB – DATA BASE”, which affects the main principles of the
linguistic organization.
2. The second stage, initiated by Sir Timothy Berners-Lee, in the form of
computer realization of this problem by using simplified speech message presen-
tation models only through a plurality of triples (Subj, Pred, Obj); O.O. Varlamov
has complicated the message presentation scheme by simultaneous use of the tri-
ples (Subj, Pred, Obj) and the logical functions “if .. 1S .. then .. nS ”, which con-
nect separate situations in the form of messages of a monopredicative level.
3. The third stage (we can assume) is presented by the author’s work (“Back
to basics of speech activity”) [8], it forms a new platform for studying the lan-
guage activty, by introducing the concept of the Basic Semantic and Syntactic
Structure (BSSS), which is derived from the structural and functional level of
neural organization of the visual pathway and which will be the same for all hu-
manity. That is, for the unit of perception, processing and “understanding” of
language material it is necessary to take a separate BSSS, which simultaneously
becomes a semantic unit of formation and presentation of an arbitrary text. The
BSSS structure was presented at Fig. 5.
Of course, the question arises as to how complex is this presentation of natu-
ral language for a computer comparing with previous versions - the arguments
here are as follows:
1) BSSS is a derivative from the structural and functional level of the neural
organization of the visual pathway - hence, it will be the same for all languages;
2) BSSS is clearly and unambiguously formalized at the structural level;
3) it is completely conscious and correctly perceived by a person;
4) it appears as a structural unit for formation of an arbitrary message;
5) if the BSSS structure can be clearly and unambiguously presented at the
linguistic level, then, apparently, in the same formalized way it can be translated
to the computer level;
6) perhaps the main thing: it is possible to form a computer technology for
processing PM texts in an automatic mode without a compulsory scheme for the
formation of metatext for the computer.
FINAL PART
Only in the process of working with the given topic (in fact, during the last final
stage), the author himself realized the clear semantic load of the correlation
between notions “Semantic Web – Data Base”.
“Semantic Web” is a semantic matrix (standard relationship system), which
we constantly use at the stages of analysis or synthesis of speech messages. This
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 138
is the level of our linguistic competence regarding the structural and functional
levels of linguistic organization that provide us with and guarantee adequate
character of implementation of the communication process in the modes of
synthesis / analysis of speech communication. Our linguistic competence is
formed, as they say “with mother’s milk” and the person / child masters the
linguistic system at about 2,5-3 years, and confidently uses it at about 10-12
years.
There has been created (synthesized) nothing better than the “matrix of
linguistic communication in the form of basic semantic and syntactic structure”
by the nature, since this scheme is derived from the structural and functional level
of the neural organization of the visual pathway; so, it is standard – the same for
all humanity. This is a generalized scheme of how a person perceives all available
information on the sensory or linguistic level, and the ability to translate it into the
language level in the form of a messages of the mono/poly- predicative level.
Data Base, in this case, is a generalized scheme of accumulation, storage
and presentation of the accumulated product by our linguistic activity for
perpetual storage. And all attempt to change, transform, improve this
communication scheme are useless. The only possible scheme of reproduction of
the human cognitive potential is the BSSS structure. We should work for the
computer to be able to understand this structural organization in the form of BSSS
sequence.
Data Base is the big cognitive potential accumulated by humanity through-
out the development of civilization, stored in papyrus, manuscripts, books, audio /
video records. That is, all efforts should be directed for the computers, networks,
distributed systems to store and generalize, but not lose the main values of our
civilization.
Well, after carefully analyzing the structural level of the linguistic organiza-
tion, we can come up with a more reasoned analysis of the individual language
system (see Fig. 1), the components of which are the interconnected powers of the
linguistic processor (LP) and the basic knowledge (BK). The first component de-
fines the structural and functional levels of organization of the language channel
as the basic scheme of perception and accumulation of the cognitive potential of
humanity at the linguistic level, while the second – the BK defines the collective
cognitive potential of a person, which stores all information both at the figurative
level (obtained through the entire sensor system), and at the language level re-
ceived through the perception of information by visual or audio channels. The
latter, upon the request of the recipient, can always be translated to the language
level.
The present study considers the functioning of LP and BK exclusively at the
linguistic level. It is appropriate to emphasize the significant assistance of post-
graduate students D.S. Serheiev and A. S. Khimicha in developing these issues.
This is basically a team of like-minded people who have fruitfully worked on this
ambitious program – modeling the speech behavior of a person in the form of an
interconnected LP-BK system. The author expresses his sincere gratitude for co-
operation in developing this area of research.
If Semantic Web is responsible for the process of perceiving and “under-
standing” the text, which, according to the integration approach, is based on a plu-
rality of standard structures, then we can assume that the cognitive element of
storage, and hence – the understanding and using the accumulated knowledge,
On the way to solving the problem of “semantic web – data base”
Системні дослідження та інформаційні технології, 2019, № 2 139
will be the basic semantic and syntactic structure itself. We should also note that
the whole cognitive potential of a person is formed from a plurality of situations
not only of symbolic, but also of figurative levels, preserving mother’s voice
(with all its shades), appearance, character, all her attributes at the neural level.
The given work reproduces the results of the infinite stages of work on the
declared “eternal” topic, and I should note the actual, or virtual attention and sup-
port that I constantly felt due to certain individuals.
In this way, I want to note the very pleasant audience and the spirit of com-
munication at the “Artificial Intelligence” seminars, initiated by A.I. Shevchenko,
the main priorities of which were: freedom of thought, freedom of imagination,
goodwill and support for beginners. The last significant work of the author, “Per-
sonalized cognitive feedback, as a powerful lever of progressive acceleration of
the development of society” was published due to support of Anatolii Ivanovych
in the journal “Artificial Intelligence”, 2018, No. 1 [10].
Meetings and communication with the member of the academy O.O. Kryshtal
were very productive. I would like to thank for acquaintance with his interesting
treatises, covering the philosophical and linguistic problems of today, for his sup-
port to my search on the neurophysiological principles of linguistic activity; for
submitting the article “Neurophysiological grounds of the structural organization
of linguistic material”, which appeared at reports of the National Academy of Sci-
ences of Ukraine, Edition 11, 2007 [6], and which further inspired the author for
the monograph “From thought to knowledge” (neurophysiological grounds), pub-
lishing house “Ukrainian Chronicle”, 2008 [7].
It is important to emphasize the importance, urgency and timeliness of crea-
tion of the BICA (Biologically Inspired Cognitive Architectures) magazine,
which is focused on studying the cognitive processes occurring in our neural net-
work. For the first time we met (as I recall) with O.V. Samsonovich in Katsiveli
at the seminar “Artificial Intelligence”; I think that from that day, he had an idea
about creating the BICA magazine.
I am very grateful for the support and presentation of my research in this
magazine: presentation of the research direction (BICA-13, Kyiv, Grandhotel),
publication of the work “To the origins of speech activity” (BICA-14) [8], support
of this research from my graduate students Yuri Kyslenko, Danylo Sergeiev –
“Cognitive architecture of speech activity and modeling thereof” (BICA-15) [9].
In this context, one should note that when the main vector of social devel-
opment (according to B.F. Porshnev) is a “progressive acceleration process”, we
can predict that “the increase in the authority and the ranking of research of this
direction will also occur under the scheme with a progressive acceleration”.
The last work, “Personalized cognitive feedback, as a powerful lever of pro-
gressive acceleration of the development of society” [10], confirms that only a
living neurosubstance of a single individual is capable of synthesizing new infor-
mation that was absent in the database. Whether it possible to perform it by IT
technology in future, is a rhetorical question. But even this is very useful for im-
proving the cognitive potential of the society. All knowledge that can be synthe-
sized by living neurosubstances of the society and translated into the language
level, may potentially become an achievement of the artificial intelligence, pre-
served and transmitted for future generations.
In this context, one should mention the words of the famous Kyiv citizen
concerning the influence of the individual on the “progressive acceleration of the
development of society”, written on the walls of his Alma Mater – the Igor Sikor-
sky Kyiv Polytechnic Institute: “
Y.I. Kislenko
ISSN 1681–6048 System Research & Information Technologies, 2019, № 2 140
The work of an individual remains the spark that drives the society even
more than collective work” (Igor Sikorsky).
LITERATURE
1. Anderson Janna. The Fate of the Semantic Web / Janna Anderson, Lee Raine. —
Pew Research Center’s Internet & American Life Project. — May 4, 2010.
2. Щерба Л.В. О трояком аспекте языковых явлений и эксперименте в языкозна-
нии. Языковая система и речевая деятельность / Л.В. Щерба. — М., 1974.
3. Кисленко Ю.І. Архітектура мови (лінгвістичне забезпечення інтелектуальних
інтегрованих систем): навч. посібник / Ю.І. Кисленко. — К.: Віпол, 1998. — 343 с.
4. Зеки Семир. Зрительный образ в сознании и мозге: сборник трудов / Семир
Зеки // В мире науки. — № 11–12. — М.: Мир, 1992.— C. 33–41.
5. Хокинс Дж. Об интеллекте / Дж. Хокинс, С. Блейксли. — М.: Издат. дом
«Вильямс», 2007. — 240 с.
6. Кисленко Ю.І. Нейрофізіологічне підґрунтя структурної організації мовного
матеріалу (Представлено академіком НАН України О.О. Кришталь) /
Ю.І. Кисленко // Доповіді НАНУ. — 2007. — №11. — C. 158–164.
7. Кисленко Ю. От мысли к знанию (нейрофизиологические основания): моногр. /
Ю. Кисленко. — К.: Издательство «Український літопис», 2008. — 102 c.
8. Kislenko Y.I. Back to basics of speech activity / Y.I. Kislenko // Biologically In-
spired Cognitive Architectures. — 2014. — Vol. 8. — P.46–68.
9. Kyslenko Yuri. Cognitive architecture of speech activity and modeling thereof /
Y.I. Kislenko // Biologically Inspired Cognitive Architectures. — 2015. —
Vol. 12. — P.134–143.
10. Kislenko Y.I. Personified cognitive feedback as a powerful instrument for progres-
sive acceleration of social evolution / Y.I. Kislenko // Штучний інтелект. —
2018. — № 1. — P.63–95.
11. Port-Royal Grammar. — 1966.
12. Ф. де Соссюр. Курс общей лингвистики. — М.,1963.
13. Пиотровский Р.Г. Лингвистические уроки машинного перевода / Р.Г. Пиот-
ровский // Вопросы языкознания. — 1985. — № 4.
14. Астахова Л.И. Предложение и его членение: прагматика, семантика, синтаксис
/ Астахова Л.И. — Днепропетровский ГУ, 1992.
15. Грамматика современного русского литературного языка. — М.: Наука, 1970.
16. Городецкий Б.Ю. Компьютерная лингвистика: моделирование языкового об-
щения / Пер. с англ. / сост., ред. и вступ. ст. Б.Ю. Городецкого) // Серия
«Новое в зарубежной лингвистике». — Вып 24. — М.: Прогресс, 1989. — 432 с.
17. Гвоздев А.Н. Формирование у ребенка грамматического строя русского языка /
А.Н. Гвоздев. — М.: Изд-во АПН, 1949.
18. Кисленко Ю.І. Структурно-функціональний рівень організації лінгвістичного
процесора / Ю.І. Кисленко, А.В. Хіміч // Системні дослідження та інформа-
ційні технології. — 2018. — № 1. — C. 19–35.
19. Ибука Масару. После трех уже поздно / Масару Ибука. — М.: Знание, 2000. — 192 с.
20. Кисленко Ю.И. Проблемы и перспективы развития поисковых систем /
Ю.И. Кисленко, А.В.Терентьев // Искусственный интеллект. — 2011. — № 3.
21. Сегаран Тоби. Программируем коллективный разум / Тоби Сегаран; пер.
с англ. — Символ-Плюс. — 368 с.
Received 24.03.2019
From the Editorial Board: the article corresponds completely to submitted
manuscript.
|
| id | journaliasakpiua-article-175555 |
| institution | System research and information technologies |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2025-07-17T10:26:05Z |
| publishDate | 2019 |
| publisher | The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" |
| record_format | ojs |
| resource_txt_mv | journaliasakpiua/ab/856548b7c86322b8c28a34f48092b2ab.pdf |
| spelling | journaliasakpiua-article-1755552019-08-27T22:12:50Z On the way to the problem of "Semantic Web – Data Base" На пути к решению проблемы "Semantic Web – Data Base" На шляху до вирішення проблеми "Semantic Web – Data Base" Kislenko, Y. I. semantic Web data Base integration approach quantum of knowledge the basic semantic-syntactic structure semantic Web data Base інтеграційний підхід квант знань базова семантико-синтаксична структура semantic Web data Base интеграционный подход квант знаний базовая семантико-синтаксическая структура The present state of the problem of "Semantic Web – Data Base" is analyzed. "Semantic Web" is analyzed from the standpoint of the integration approach, covering the results of research in the fields of neurophysiology, psychology, philosophy; it allows to formally define the quantum of knowledge as a separate situation of the visual level and clearly define the scheme of its verbalization in the form of a basic semantic-syntactic structure. The main result of the integration approach is presented by the following thesis: “The structural level of the linguistic organization is derived from the structural and functional level of the neural organization of the visual path”. From here we have a productive conclusion: the structural and functional level of the language organization will be the same for all languages. The second component of the "Data Base" covers (should cover) the entire social cognitive potential of knowledge, presented by a plurality of accumulated texts. The structured level of knowledge base organization is presented by a very small fragment of the neural network, which reproduces a separate situation of text information, but which, through a plurality of separate tokens, of its constituents (with corresponding references to other structural formations) forms a practically cognitive neural network of a certain knowledge area. Проанализировано современное состояние решения проблемы "Semantic Web – Data Base". Анализ "Semantic Web" выполнен с позиций интеграционного подхода, охватывающего результаты исследований в сферах нейрофизиологии, психологии, философии; это позволяет формально определить квант знаний как отдельную ситуацию зрительного уровня и четко очертить схему ее вербализации в виде базовой семантико-синтаксической структуры. Главный итог интеграционного подхода презентован тезисом: "Структурный уровень организации языка является производным от структурно-функционального уровня нейроорганизации зрительного тракта". Отсюда следует важный вывод: структурно-функциональный уровень языковой организации будет одинаковым для всех языков. Составляющая "Data Base" охватывает (должна охватывать) весь общественный когнитивный потенциал знаний, представленный множеством накопленных текстов. Структурный уровень организации базы знаний презентованный небольшим фрагментом нейронной сети, который воспроизводит отдельную ситуацию текстовой информации, но который через множество отдельных лексем (с соответствующими ссылками на другие структурные образования) формирует практически когнитивную нейросеть определенной области знаний. Проаналізовано сучасний стан вирішення проблеми "Semantic Web – Data Base". Аналіз "Semantic Web" виконано з позицій інтеграційного підходу, що охоплює результати досліджень у сферах нейрофізіології, психології, філософії; це дозволяє формально визначити квант знань як окрему ситуацію зорового рівня та чітко окреслити схему її вербалізації у вигляді базової семантико-синтаксичної структури. Головний підсумок інтеграційного підходу репрезентовано тезою: "Структурний рівень мовної організації є похідним від структурно-функціонального рівня нейроорганізації зорового тракту". Звідси випливає продуктивний висновок: структурно-функціональний рівень мовної організації буде однаковим для всіх мов. Складова "Data Base" охоплює (повинна охоплювати) весь суспільний когнітивний потенціал знань, презентований множиною нагромаджених текстів. Структурний рівень організації бази знань презентовано невеликим фрагментом нейронної мережі, який відтворює окрему ситуацію текстової інформації, але який через множину окремих лексем її складових (з відповідними посиланнями на інші структурні утворення ) формує когнітивну нейромережу певної галузі знань. The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" 2019-06-25 Article Article application/pdf https://journal.iasa.kpi.ua/article/view/175555 10.20535/SRIT.2308-8893.2019.2.10 System research and information technologies; No. 2 (2019); 115-140 Системные исследования и информационные технологии; № 2 (2019); 115-140 Системні дослідження та інформаційні технології; № 2 (2019); 115-140 2308-8893 1681-6048 en https://journal.iasa.kpi.ua/article/view/175555/175468 Copyright (c) 2021 System research and information technologies |
| spellingShingle | semantic Web data Base інтеграційний підхід квант знань базова семантико-синтаксична структура Kislenko, Y. I. На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title | На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title_alt | On the way to the problem of "Semantic Web – Data Base" На пути к решению проблемы "Semantic Web – Data Base" |
| title_full | На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title_fullStr | На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title_full_unstemmed | На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title_short | На шляху до вирішення проблеми "Semantic Web – Data Base" |
| title_sort | на шляху до вирішення проблеми "semantic web – data base" |
| topic | semantic Web data Base інтеграційний підхід квант знань базова семантико-синтаксична структура |
| topic_facet | semantic Web data Base integration approach quantum of knowledge the basic semantic-syntactic structure semantic Web data Base інтеграційний підхід квант знань базова семантико-синтаксична структура semantic Web data Base интеграционный подход квант знаний базовая семантико-синтаксическая структура |
| url | https://journal.iasa.kpi.ua/article/view/175555 |
| work_keys_str_mv | AT kislenkoyi onthewaytotheproblemofquotsemanticwebdatabasequot AT kislenkoyi naputikrešeniûproblemyquotsemanticwebdatabasequot AT kislenkoyi našlâhudoviríšennâproblemiquotsemanticwebdatabasequot |