=Paper=
{{Paper
|id=Vol-2416/paper48
|storemode=property
|title=Multimodel clustering of social networks in social dampening applying BIG DATA (acquiring knowledge from data)
|pdfUrl=https://ceur-ws.org/Vol-2416/paper48.pdf
|volume=Vol-2416
|authors=Irina Khaimovich,Vladimir Ramzaev,Vadim Chumak
}}
==Multimodel clustering of social networks in social dampening applying BIG DATA (acquiring knowledge from data)==
https://ceur-ws.org/Vol-2416/paper48.pdf
Multimodel clustering of social networks in social dampening
applying BIG DATA (acquiring knowledge from data)
I N Khaimovich1,2, V M Ramzaev1 and V G Chumak1
1
Samara University of Public Administration “International Market Institute”, 21,
G.S.Aksakova Street, Samara, Russia 443030
2
Samara National Research University, Moskovskoe Shosse 34А, Samara, Russia, 443086
e-mail: kovalek68@mail.ru
Abstract. The developed methodology provides a solution to two essential tasks, thereby
revealing the gnoseological potential of Big Data technology: social forecasting in the three
most significant areas of the information society based on a model which identifies conditions
for social resonance; successful implementation of the social dampening procedure based on
the use of appropriate management options using multimodal clusterization of social networks
based on Big Data technology. The article suggests the tool that helps to increase work
efficiency in the sphere of social dampening in the region. The proposed method of regulation
may be efficient when it comes to the control of the regional social dampening processes which
have variety of forms and broad range of elements and factors, as well as growth dynamics and
active transformation of life activities. At the same time using modern products make it
possible to evaluate and show changes on a real-time basis which can be useful for local
government authorities.
1. Introduction
Let us consider the analysis of the gnoseological potential of Big Data from the standpoint of
synergetics, which focuses mainly on unbalanced, disordered systems, which are formed both in
natural and social environments, which acquire balance only at certain moments, often extremely
fleeting, but no less important and requiring comprehensive thinking.
Certainly, the destabilization of the system, and above all the social system, is extremely interesting
from the dialectical positions both as awareness of the reasons, and in terms of exploring the
possibility of providing the system with a steady state. However, the pragmatic reality convinces us
that numerous ordinary people who are the direct elements of various systems, as well as the political
elite, who seek to hold the position of a leading social force possessing certain levers of influence on
social dynamics, are still primarily interested in the stability of the system structures. Deviation from
the average values of social indicators is clearly perceived as deviance. Under these conditions, the
technical capabilities of Big Data allow us to consider them as a tool for indicating the level of
deviation of the most diverse social processes from the optimal model for a given society, contributing
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019)
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
to the development of a mechanism for returning to the averaged, balanced form of both a single
process and their complex.
In our opinion, in order to conduct a scientific analysis of social processes, it is necessary to reveal
the essence of the many-sided deviations that arise in certain situations in society. The analysis carried
out in this direction shows the possibility of identifying two main types of changes in social processes.
The first type is characterized by a rather slow deviation of social processes from the norms and
principles established in society, described by an additive function. In this case, social smoothing,
response can be considered as a cognitive impact on each social subject in particular, regardless of its
nature: individual or collective. Here, the dominance of the elementary approach to the description of
reality manifests itself, in which the quality of a system is determined by the quality of its constituent
elements, which are quite accessible for comprehensive analysis using Big Data.
The essence of the second type of changes in social processes is determined by the explosive
changes inherent in it and, first of all, by social cataclysms that bring society out of balance. The
second type in its essence reveals deviations of the multiplicative type, which are intensified due to the
extremely low efficiency of the interaction processes of social subjects. Here, in contrast to the first
type, a social explosion occurs in a sharper form, much more concentrated in time and determined in
many respects by the disadvantages of intergroup interaction, similar to social resonance. This
situation is an example of the manifestation of a systematic approach to reality in social practice,
based on the principles of which the quality and, therefore, the sustainability of a social system
depends on the links of its constituent elements.
In order to analyze social processes classified in the above manner, ensure their management and
bring social deviations to an acceptable standard, the authors propose to introduce into the conceptual
apparatus the term “social dampening”, fundamentally new to social and humanitarian knowledge.
This concept is intended to denote the desire of society to average in terms of the manifestation of its
activity, therefore, in stability. It is expedient to characterize such stable existence of a society as a
social equilibrium, a special state of a social system that has developed due to historical continuity,
which manifests itself in a whole set of factors of economic, social, ethnic, etc. character.
Deviation from the criteria of social stability is the essence of social balance and leads to a “lifting”
of the social equilibrium curve, which necessitates social dampening as a significant manifestation of
management processes that can, among other things, solve the problem of cognitive control of
consciousness, differing in its mental manifestations.
“Leveling” of quantitative (digital) characteristics should be carried out in indicators acceptable
for a given society. Using Big Data as the most sensitive tool for measuring and detailing biogeosocial
processes allows for the necessary iterations that cannot be performed using conventional scientific
tools, identifying the “dominant that directs the vector of development of a specific phenomenon or
process” [1] and the trend of their changes.
The formulation of the problems presented by the authors of the study is also aimed at clarifying
the issue related to the need to determine the vector of the social management algorithm itself, namely:
if we set a completely obvious goal for us to ensure proper management of social systems, then the
following question arises: what (as a social goal) should we control in this way. The use of Big Data
technologies provides new opportunities for solving such problems. This, in turn, means the need to
determine the factors and criteria for the stable state of structures, and more precisely, the control
objects, in the system of social coordinates. The frequency of deviations from these indicators, as well
as the ability of their leveling, according to the authors, will be a fundamental point in understanding
the quality of social management, which actualizes the need to develop a methodology for assessing
the degree of deviation of social processes and the qualitative state of the systems involved in them
from their equilibrium states. This method should necessarily include the possibility of both direct and
indirect impact on social systems, which, in fact, additionally reveals the above concept of “social
dampening” as a fundamentally significant theoretical basis for the formation and subsequent
implementation of managerial influences on society, allowing for reduce, by using Big Data, the peak
values of social indicators that go beyond the social acceptability. If we ignore them it can not only
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 377
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
partially unbalance the social system, but make it fully socially uncontrollable and prone to social
explosions.
At the same time, in addition to social processes leading to a deviation of the environment from the
social “normal”, which can be regulated by social dampening, it is necessary to consider in greater
detail identifying the essence of the process of social resonance. Social resonance is such a state of
interacting factors and criteria that characterize the social environment, which leads to an explosive
deviation from the “normal”. The “removal” of such social aggravations can be interpreted as a
process of returning from social resonance to the norm, which seems to the authors an extremely
important task of social management, the successful solution of which is determined by the need to
take into account the effect of passionate interaction, understood as a kind of “activity manifested in
the individual’s striving for the goal (often - illusory) and in the capacity for superstressing and
sacrifice in order to achieve this goal”, at the same time “sacrifice is understood ... as refusal to satisfy
immediate needs, sometimes essential to life, for the sake of the dominant social or ideal needs,
perceived as a goal “with a predominance of “development needs”[1].
Thus according to the authors, Big Data makes it possible to form a model of “removing” of social
deviations, including social resonance, arising from a combination of a wide range of factors, the
interaction of which is very difficult to analyze and often invisible to the researcher through the prism
of traditional cognitive tools. In its turn, the implementation in practice of this kind of social
dampening model let us create a comprehensive understanding of social processes as a neurobiological
manifestation of human activity and forms the necessary opportunities to overcome its negative
consequences.
Hiding the analysis of additive and multiplicative effects behind the facade of large-scale computer
calculations, Big Data makes it possible to ensure that objective decisions are made, especially
significant in the management of multifactorial social and natural systems [2,3], which, thanks to the
implementation of social dampening procedures, makes it possible to guarantee stability,
manageability and predictability of biogeosocial processes.
The carried out research clearly convince that the methodology of this kind must consist of two
semantic blocks: the methodological, structuring algorithm for the study of social systems and
processes with Big Data technology and management, which determines the possible methods of
management influence based on the carried out analysis.
The first methodological block supposes, first of all, isolation, accompanied by analysis and
evaluation, in the sphere of the information space of multimodal clusters, the most explosive in terms
of their potential to destabilize the social system, taking it out of relative equilibrium. Here, as global
clusters, attention is drawn to the three classical spheres of society: political, economic and social. The
undeniable scale of these spheres makes it possible to isolate subclusters in each of them with the
possibility of further detailing them into groups (including interests) and IP addresses.
This specification is quite possible to carry out on the basis of keywords that are set in accordance
with the research task and are used by the Big Data technology to isolate subclusters and elements of
their internal structure. Further, the practical implementation of this stage of the proposed
methodology will require the parallel development of the “system learning” algorithm, allowing the
program to isolate words, terms, concepts, etc. for the subsequent clustering and identification of
significant communications.
The next step of the first methodological block is the construction of an algebraic lattice of the
number of links of multimodal clustering, which represents a peculiar coordinate system of the
specific carried out analysis. It will allow modeling the system of links within the cluster. The increase
in the number of links within the cluster (or its subsystem) is the main indicator of the subsequent
system out of equilibrium - social resonance.
The greatest interest for the subsequent management impact should present connections that
demonstrate sustainable growth. When they are identified, you should:
- precise the topics that ensure the applicability of these links;
- carry out an analysis of the frequency dynamics;
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 378
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
- identify danger zones for social resonance.
The completion of the first methodological block is the differentiation of the resonance rate. This
involves the identification of areas of social, informational interactions on which there has been a
steady growth; subsequent finding of stable relationships to identify the topic of interaction. At the
same time, a sharp increase in the number of links per unit of time by an order of magnitude should be
interpreted, within the framework of this methodology, as the main condition for the soonest
occurrence of social resonance.
The second methodological block identifies possible options for a practical transition from
gnoseological reasoning to direct management impact on the social situation. A management decision
point as a manifestation of social dampening can be represented on the graph at the point of transition
from a smoothly increasing curve, indicating an increase in the number of social connections (within
the identified cluster, subcluster, etc.) to a vertical straight line, representing the ordinal increase in
social connections.
In this case, the second methodological block admits the possibility of two management impacts:
- Management of the number of socially significant connections, which primarily involves
minimizing them in order to prevent or level social resonance;
- Management of the social environment dissipativity, which provides for an instrumental,
informational impact (including through informational attack or subject readdressing of links) on the
identified cluster, an impact that does not allow the formation of social relations that could in the long
run lead to social resonance.
Otherwise, the coincidence of the “resonance frequencies” in the process of intersubject interaction
can cause a global, uncontrolled social resonance.
2. Application of the formal concepts analysis method for social networks
The methodology of social dampening consists of the social resonance definition in social networks
groups and the search for control actions to reduce tensions. The methodology of social dampening is
based on multimodal clustering of social networks. Clustering is based on the formal concept analysis
method (Formal Concept Analysis, FCA) [4–9]. A large amount of structured and unstructured data
generates trivial data.
According to the method of formal concepts, we introduce the following definitions:
G is a set of objects, M is a set of attributes, relation I G M such that ( g ,m ) I if and only if,
the object g has attribute m. K : ( G , M , I ) is called a formal context.
Let us define Galois operators in the following manner for A G, B M :
def def
A' { m M g / mg A }, B' { g G g / mm B } .
A formal concept is the pair ( A, B ) : A G, B M , A' B and B' A , where A is a formal extent, В
is a formal intent.
The concepts, ordered by the dependency ( A1 , B1 ) ( A2 , B2 ) A1 A2 ( B2 B1 ) , form a complete
lattice, which is called context lattice ( G , M , I ).
The example of social networks context and their context hashtags is shown in Figure 1.
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 379
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
a)
b)
c) d)
Figure 1. Social networks with context hashtags: A) context hashtags for social networks; B) social
hashtags for social networks with implications; C) hashtags with reliability of associative links; D) the
example of hashtags with support of association rules.
Let us consider implications in the lattice and define the operation “implication” in the following
manner: implication A B , where A, B M exists if A' B' , i.e. every object that has all attributes
from A set also has all attributes from B set. Implications correspond to Armstrong’s axioms:
reflexivity ( A B / A C B ) ;augmentation ( A B / A C B ) ; pseudo-transitivity
( A B, D B C / A C ) .
Among implications there is the Duquenne-Guigues basis of implications i.e. a minimal number of
implications which can help to derive other implications using the Armstrong rules. The basis is
sought through methods of machine learning: “object-by-object” algorithm of implication basis
construction or through interactive learning procedure. As a result we have a hashtag with implications
(Figure 2б).
Implications: abc d ,b c,cd b allow to redefine the nodes of hashtags.
Let us consider partial implications or association rules.
Definition 1. A m
,n
B is partial implication (association rule) of context (G,M,I), if A, B M ;
has a support of sup p( A B ) ( A B )' / G has a confidence conf ( A B ) ( A B )' / A' .
Further we consider the algorithm identifying association rules:
1) to find all frequent sets of attributes (with support which is not below the assigned);
2) it is enough to find all the frequent closed sets of attributes of the context.
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 380
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
The example of hashtags with reliability of associative links (Figure 1c).
The example of hashtags with support of association rules (Figure 1d).
Good rules with sup p 1/ 2 and min conf 3 / 4 are defined according to the following algorithm:
1. 0 c, sup p( 0 c ) conf ( 0 c ) 3 / 4;
2. c b, sup p( c b ) 1 / 2,conf ( c b ) 2 / 3.
The use of this clustering method will allow determining the interest groups, with an increase in
connections it will be necessary to make management decisions. In terms of information, it will be
necessary to combine messages from groups with an increased number of links with calm “stable”
groups in which the number of links does not undergo a drastic change.
3. Method of Using Data Mining in Social Dampening Methodology
Formal Concept Analysis (FCA) allows to establish stable links and cluster data (create new
knowledge) using Armstrong rules in context lattices. To create a portrait and information model of
resonant groups, it is necessary to use BIG DATA technologies [10, 11, 12]. The method of using data
mining is as follows:
1. Forming a big data set in hadoop from twitter using the “Samara region” filter which reveals hit
counts;
2. Separation of the formed set by various filters associated with the basic factors of resonance
deviations;
3. Monitoring of streaming content analysis on filters;
4. Adoption of operational activities in cases of sustained “hits” in hit counts;
5. Development of a program in the Scala language for working with filtering in Big Data field;
6. Debugging and testing the program with a set of practical data;
7. Analysis of the calculation results.
The social network twitter is used to obtain data, since it is an “open” product, its use does not
require additional investments, and 50% of Internet users have profiles in this program. Twitter is the
second most popular network among users worldwide, second only to Facebook. However, unlike
Facebook, which does not provide open access to its data, Twitter provides such access, there are no
restrictions on access to server data sets. Users of this social network exchange mostly text
information, which is an undoubted advantage in processing. Twitter is not a subject network and most
widely reflects public opinion on many issues of interest, so for the formation of groups for analyzing
the social resonance in the region, the processing of data from this social network was optimal.
To work with BIG DATA in social networks, it is necessary to use methods of collecting,
processing and analyzing data. Data collection is carried out in real time, within a certain geolocation,
or within the entire network, using certain patterns. Information of interest for the analysis in this area
is: location, date and time, content, content “author” (user), communication between users. Data
collection in social networks can be performed using the following tools: Apache Hadoop, Biglnsights
(IBM), Cloudera, Hortonworks, Storm. Hortonworks was chosen to carry out research in the field of
social dampening. The Twitter Application (apps.twitter.com) was used for work, in which key
parameters were defined and refined: API key, API secret, Access token, Access token secret.
To collect data using the Hortonworks, Twitter App, the flume service configuration file was used
in the Hortonworks Sandbox virtual machine. After installing the Hortonworks_ Sandbox version 2.3
virtual machine and setting up the flume service, the system is ready to download data from twitter. To
view and download the downloaded files, go to the HDFS folder, where we process the data. View of
the HDFS file structure in the Hortonworks virtual machine when solving a problem in the sphere of
social dampening is shown in Figure 2.
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 381
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
Figure 2. HDFS visualization in Hortoworks when file download to solve social dampening.
The collected data must be structured (i.e., processed) in accordance with the MapReduce
paradigm. MapReduce is a framework for performing distributed tasks using a large number of
computers forming a cluster.
Using MapReduce allowed us to structure the flow of data from social networks by criteria: fonts,
text size, color, link to user profile, location, time, and so on.
To determine the portrait of the respondent, the following types of data are needed: location, text,
language, and time. In order to extract only this information, you can use the MapReduce technology
built into the Hortonworks Sandbox tool. For data processing, we use the Hive DBMS in the Hadoop
environment, which allows performing operations on data and their analysis using SQL-like queries.
To do this, create a file for processing and creating the necessary hivedll.sql tables.
Run this file using command: Hive_f hiveddl.sql. Structured data will be presented in the Table 1.
Table 1. Headings to analyze structured XML data in tasks for social dampening.
A B C D E F
Data/Time Time/Zona language Text location Sentiments
For data analysis the following variables are used. Total amount of twitts (Koli) for every location
(R) is defined by:
N
KolR ki , ki R,
i 1
where ki is every next twitt in the considered stream.
Frequency of unique word usage ch(m) is defined from the general variety of L text data:
N
ch(m) mi , mi L.
i 1
The attitude of every twitt otn (m,rez) may be defined from the thesaurus tez, where the attitude to
this word is written up:
0, m negative _ meaning
otn( m , rez ) 1, m neural _ meaning .
2, m positive _ meaning
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 382
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
For further work, a dictionary was compiled, consisting of domain filters, to further determine the
number of tweets for placement ch (m) and the number of tweets for placement taking into account the
relation otn (m, rez). We define a thesaurus taking into account filters by basic factors: salary,
unemployment, political developments and housing and utility infrastructure. As a result, we obtain 4
basic factors of resonance deviations.
«Salary» factor P1 gives the number of twitts in general quantity of text data L:
N
KolotnP S i ( S i P1 ) / L 10% .
1 i 1
«Unemployment» factor P2 gives the number of twitts in general quantity of text data L:
N
KolotnP2 S i ( S i P2 ) / L 9% .
i 1
«Political» factor P3 gives the number of twitts in general quantity of text data L:
N
KolotnP3 S i ( S i P3 ) / L 7% .
i 1
«Infrastructure» factor P4 gives the number of twitts in general quantity of text data L:
N
KolotnP4 S i ( S i P4 ) / L 13% .
i 1
4. Results and discussion
As a result, it can be concluded that what factors of deviations are associated with the Samara region.
According to Figure 3 it can be seen that the main factor of discontent is the housing sector.
Figure 3. Factors of social dampening in Samara Oblast.
Due to BIG DATA technology, it is possible to store and update data in the “hаdoop” file system
using the “Samara Oblast” filter (filter1 = {Samara Oblast}). Then, it is necessary to filter this area by
the basic factors of social dampening, by installing, for example, the following filters: Filter2 (salary)
= {money, ruble *, dollars *, currencies *, crypt *}; Filter3 (unemployment) = {job search, engineer *,
worker *, build *}; Filter4 (political developments) = {elections, deputy *, penny *, administrator *};
Filter5 (housing and utility infrastructure) = {garbage collection, pipes *, water *, gas *}.
The set of descriptors by which the Internet discourse will be filtered is determined by the lexical
representatives of the concept formed in the world picture by the average Russian-speaking consumer.
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 383
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
To make decisions in the field of social resonance in the region, multimodal clustering of social
networks was carried out. A large number of structured and unstructured data of social sites in the
considered area can be represented as the next triple (user, group, interest) (Figure 4).
Figure 4. Data for social resonance analysis from Twitter social network as a graph
Using the method of formal concepts, we form a complete lattice, called the context lattice
(G, M , I ). An example of the context of social networks in the field of social dampening and their
context hashtags are shown in Table 2 and Figure 5.
Table 2. The example of Social network context and its context hashtags (a - attributes on
«salary» filter, b - attributes on «unemployment» filter, c - attributes on «political developments»
filter, d - attributes on «housing and utility infrastructure» filter).
G/M a b c d
1 pensioners x x
2 employees x x
3 workers x x
4 students x x x
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 384
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
Figure 5. Context hashtags for social network.
The use of this clustering method will give the opportunity to determine interest groups, with an
increase in links in which management decisions will be required. But this tool has limitations on use.
Users who work with the social network “Twitter” are in the “students” group and partly in the
“employees”, “workers” groups and only slightly affect the “retirees” group, so for complete
management decision making it is necessary to add new groups.
You can get graphs of the user hit counts by filters on data collection time (Figure 6). The data
collection time from the Internet is unlimited in BIG DATA technology.
Figure 6. Dependency graph of hit counts using filters from data collection time.
As a result, we obtain a dynamic change of information in real-time from the Internet, which allows
monitoring the streaming analysis of unstructured information (In-Memory Data Processing and
Stream technology) with minimal investment by filters. To implement this method, a program in the
Scala language was written.
After the work of the program, we obtain a dynamic change of parameters in the BIG DATA
environment, which allow us to determine social resonance zones in the region, taking into account
unstructured information. If steady “bursts” of data on hits counts are detected on charts in accordance
with the forms of resonance, management regulation for this type of activity in the region should be
implemented.
Thus, a tool has been proposed to increase the effectiveness of work in the field of social
dampening in the region. This is the most important task under modern economic conditions, the basis
of which is the possibility to make optimal management decisions. The proposed method of regulation
can be effective in managing the processes of social dampening of a region, which are characterized
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 385
Data Science
I N Khaimovich, V M Ramzaev and V G Chumak
by a variety of forms and a wide range of components and factors, as well as an inherent dynamics of
development and active transformation of life activity.
At the same time, the use of modern software and hardware gives the opportunity to make the
assessment and visualization of changes in fact in real time, which can be useful for local authorities.
It is important to note that social dampening is not a strict limiter of social actions “in amplitude”. It
only softens social actions, allowing them to manifest themselves in other areas, it does not “break”
the social system, but allows it to transform, creating the visibility of smooth compliance with the
requirements of the social interaction subjects.
5. References
[1] Bodrov A A, Ramzaev V M 2015 Philosophic aspects of developing new knowledge under
data intellectual analysis (BIGDATA) CEUR Workshop Proceedings 1490 338-345
[2] Khaimovich I N, Ramzaev V M and Chumak V G 2016 Use of big data technology in public and
municipal management CEUR Workshop Proceedings 1638 864-872
[3] Khaimovich I N, Ramzaev V M and Chumak V G 2015 Challenges of data access in economic
research based on Big Data technology CEUR Workshop Proceedings 1490 327-337
[4] Wille R 1982 Restructuring lattice theory: An approach based on hierarchies of concepts Ordered
Sets. NATO Advanced Study Institutes Series (Series C – Mathematical and Physical Sciences) 83 p
470
[5] Ganter B, Whille R 1996 Formale Begriffsanalyse (Springer, Heidelberg) p 540
[6] Ganter B, Whille R 1999 Formale Consept Analysis (Springer) p269
[7] Davey B, Priestly H 1990 Introduction to Order and Lattices (Cambridge University Press,
Cambridge) p 460
[8] Denecke K, Erne M and Wismath S L 2004 Galois Connections and Applications (SpringerScience
and Business Media) 565 p 498
[9] Bonacich P 2007 Power and Centrality: A Family of Measures American Journal of Sociology
92(5) 1170-1182
[10] Сhumak P V, Ramzaev V M and Khaimovich I N 2015 Models for forecasting the competitive
growth of enterprises due to energymodernization Studies on Russian Economic Development 26(1)
49-54
[11] Khaimovich A I, Grechnikov F V 2015 Development of the requirements template for
the information support system inthe context of developing new materials involving Big Data
CEUR Workshop Proceedings 1490 364-375
[12] Kazanskiy N L, Stepanenko I S, Khaimovich A I, Kravchenko S V, Byzov E V and Moiseev M
A 2016 Injectional multilens molding parameters optimization Computer Optics 40(2) 203-214 DOI:
10.18287/ 2412-6179-2016-40-2-203-214
V International Conference on "Information Technology and Nanotechnology" (ITNT-2019) 386