=Paper=
{{Paper
|id=Vol-2763/CPT2020_paper_s4-4
|storemode=property
|title=The software platform for creating and conducting artificial intelligence competitions with a visualization subsystem
|pdfUrl=https://ceur-ws.org/Vol-2763/CPT2020_paper_s4-4.pdf
|volume=Vol-2763
|authors=Alena Zakharova,Nikita Silchenko,Rostislav Krylov,Vladimir Averchenkov
}}
==The software platform for creating and conducting artificial intelligence competitions with a visualization subsystem==
https://ceur-ws.org/Vol-2763/CPT2020_paper_s4-4.pdf
The software platform for creating and conducting artificial intelligence
competitions with a visualization subsystem
A.A. Zakharova, N.S. Silchenko, R.A. Krylov, V.I. Averchenkov
zaa@tu-bryansk.ru|silchenko.nk@gmail.com|oktopy@gmail.com|aver@tu-bryansk.ru
Bryansk state technical university, Bryansk, Russia
The article describes the solution to the problem of teaching programming skills using modern techniques and additional software.
It is proposed to use simulation modeling of interaction developed by users as part of the training of intellectual agents in a competitive
form that implement various algorithms as the main approach to the solution. Intellectual agents are presented in the form of artificial
intelligence developed by platform users, which interacts with other intellectual agents following the rules of the developed scenario for
the competition. Scenarios provide a set of capabilities for intelligent agents, an interaction environment, and a set of constraints that
participants follow. To support this simulation, it is proposed to use a specialized software platform. The platform allows organizers to
develop scenarios with a unique set of rules, and additional platform tools speed up development and allow organizers to implement
visual display to users that can be used to show competitive process. A set of built-in platform tools allows organizers to focus directly
on the rules of the competition, since the platform provides communication with participants and additional tools for calculating the
results of the competition. In additional there is a set of basic competition systems on the platform. However, if necessary, the organizers
can present their own competition format and implement it separately. The article describes the developed platform for teaching and
holding competitions in artificial intelligence. The article also examines a number of scenarios and intelligent agents.
Keywords: artificial intelligence, intellectual agent, artificial intelligence competitions.
1. Introduction 2. Analogues
At present, technologies are developing at tremendous Currently, there are platforms aimed at artificial
speed. One of the most promising areas today is the intelligence competitions among user’s intellectual agents.
development of methods and technologies based on Among the most famous, the following platforms can be
artificial intelligence. Artificial intelligence permeate distinguished: Russian AI Cup, Mini AI Cup and Google
deeper and deeper into various spheres of human life, both AI Challenge [3-5]. In the process of participating in the
in the field of science and industry, and in everyday life. competition, users are provided with additional
In addition, the training of IT specialists who are able to information about working with the platform, and a set of
create new and maintain software products, analyze big additional training materials is created for participants.
data, administer complex technical systems and etc. is However, these platforms can hardly be used to direct
especially relevant. training for the following reasons.
The acquisition of new professional competencies in a 1. Closure. Users are provided with limited functionality
dynamically changing world requires the creation and and are not allowed to implement their own scenarios.
application of new technologies in the learning process. As a result, it is not possible to build a training
Currently, there are different software platforms. program based on them.
However, these are specialized platforms for 2. Narrow focus. Focus on a fixed set of subject areas.
implementing either educational programs or competency Often, for each subject area, a new separate platform
assessment activities (competitions, hackathons, etc.), is created, completely duplicating the functionality of
including in remote form. existing platforms, making only minor changes.
For the quick creation of unique training tracks 3. Lack of control. There is no ranking of users by levels
(educational and methodological support adapted to the of knowledge and skills. All users are forced to be in
individual learning path) a tools are needed. Current trends the same group: both novice developers and
are also aimed at the gamification of the learning process professionals.
in order to improve its quality, and the competitive aspect 4. The same type of competition system. There are a
is much better motivates for independent study of huge number of competition systems, however,
materials [1]. It is important to implement active modern platforms often use the most popular or basic
pedagogical methods involving the student in the design approaches (for example FFA, double elimination,
of his own educational path, increasing his degree of etc.).
motivation for the learning process. 5. Completeness. The platforms do not provide the
There is the introduction of meta-design into the opportunity to conduct a series of competitions with
educational process [2], which allows you to organize a subsequent monitoring in order to collect statistics
space for effective education for the required and evaluate the results. Support for the competition’s
competencies, taking into account the individual scenario is over.
characteristics of students. 6. Decentralization of information. There is no
In that case, it can be argued that the creation of a new supporting methodological assistance and additional
educational platform that is adaptable and implements the information materials. The organizers provide only
above-mentioned modern approaches to education is materials directly related to the competition.
relevant. Given the above, we note that there are tools for
creating and conducting competitions with a number of
significant drawbacks. Modern technologies are
Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY
4.0)
developing at a rapid pace, new techniques, approaches to sources and to develop much faster in the direction
learning and tools that can implement them appear. Based they are interested in.
on the existing shortcomings, it was concluded that there The platform under development is based on openness.
is a need to develop a universal educational platform on Openness implies that any user of the platform can be not
artificial intelligence with the following basic properties: only a participant, but also the creator of the educational
1. Openness. The users should be provided with the full program and the organizer of the competition. The
functionality of the platform, which will allow not platform also has an ever-expanding knowledge base
only to participate in the competition, but also to act necessary for working with intellectual agents, which is
as the organizer of the competition, as well as use the also developed at the expense of platform users, through
platform in training. which the platform is also excellent in educational
2. Broad focus. The platform should be universal, which activities. The presence of high-quality and verified
will allow it to be used in training in any IT field material will allow faster to find high-quality information
(using additional materials and methodological needed in the subject area of interest [7-8], and as a result
support), to conduct competitions and test algorithms to get better information than in open sources.
in the user-developed scenarios [6]. Moreover, the platform allows you to create your own
3. Flexible control. The platform should be able to rank scenarios and visually display them [9]. At the moment, to
users according to various criteria. Both basic criteria organize a competition, the user needs to develop a
(level of skills and competencies, age), and any other software product that implements the logic of the
criteria defined by the scenario developers can be competition and establish communication between its
used. participants. This platform already has the basic
4. The flexibility of the settings. The platform should mechanisms for that, and the organizer can focus only on
contain not only the basic conducting systems built conditions of the competition and the tasks of the scenario.
into it by default, but also allow users to implement Separately, it is worth noting that platform users can be
their own, if this is necessary. united into a group and learn how to develop in
5. Renewability. This property allows you to collect competitions with each other among opponents with equal
statistics and can serve as a universal tool for knowledge.
assessing the knowledge of participants. This is
especially useful for monitoring the overall level of 3. Description of the approach to building a
educational skills. platform for modeling the interaction of
6. Centralized information. The presence of additional intellectual agents
training material on the platform on various aspects of To solve this problem, the following architecture is
programming will allow new users to always have proposed (Fig. 1).
high-quality and up-to-date information from trusted
Fig. 1. Software platform architecture
Platform users are represented as scenario developers, The platform allows you to develop in several
organizers and participants. It means that the same person directions at once, since it is universal and supports the
or group of people can act as the developer of scenario and development of scenarios, both competitive and training,
organizer of the competition. and the complexity of such scenarios depends only on the
Organizers and scenario developers work in the control task set by the authors. The module for providing a
module, and participants interact with the participant conductive system is also flexibly configured, which
module. allows you to choose a conductive system from those
In the control module, organizer configure the already implemented on the platform, or implement your
competition parameters and load the scenario. own.
The participant module contains both general training There is also a module that allows students to get
materials and specific data for the scenario. The methodological and informational software for training.
competition module and the visualization module are The module for working with educational material
interconnected. Competitions using the module can used allows you to centrally store information and use it in the
to tests by the developer, and if the competitions are learning process. At the same time, any competition can
presented in a graphical representation, the user can see the be resumed for a specific local group, which corresponds
results in a visualization module. to the renewability property.
The competition is held on the server, which is The platform has a built-in visualization module,
necessary for data protection. The server has a module for which can significantly reduce the time for developing a
mathematical calculations and a module for implementing visual display. Users can use both the tools built into the
the system. The first module is responsible for the outcome platform and, if necessary, connect their own.
of the competition, and the second for its format. Then the
data gets into the database, and from there the organizers 4. Description and testing of developed
and participants can get it. scenarios
Administrators provide technical support and On the developed platform, three test competitive
resources for computing. scenarios were created: “Sea battle”, “Snake” and “Virus
The platform implements all the principles described war”. Each scenario has its own set of rules.
above. Openness is achieved due to the lack of a global Sea battle (Fig. 2) is presented in its classic, desktop
separation of roles. Any user can be either a participant or version, with the possibility of the initial arrangement of
an organizer of one or several competitions at the same ships and directly the game process.
time. And also there is no local separation, which provides
flexibility in management.
Fig. 2. Sea battle
The snake (Fig. 3) is presented in the form of a game to be the last among those who have reached a dead end
for four participants, where the main goal of all agents is created by the enemy's snake or natural obstacles.
Fig. 3. Snake
The virus war (Fig. 4) is presented in the form of if the agent attacks the cell of another agent, it becomes
capturing a field with special conditions: an agent can impassable and free.
capture only cells adjacent to those already captured, and
Fig. 4. Virus war
The platform also supports custom single-user development, the user agent is tested by the platform using
scenarios. The idea is that the user receives a task to various input data to verify its operation. For testing, two
implement a specific algorithm. At the end of
task scenarios were created: matrix transposition (Fig. 5)
and array sorting (Fig. 6).
Fig. 5. Matrix transposition
Fig. 6. Array sorting
A few intellectual agents have been developed for all As an example of implementation, a number of
scenarios. They solve the problem in various ways. Series scenarios were demonstrated, including “sea battle”,
of tests were carried out. “snake”, “virus war”, “array sorting” and “matrix
For competition scenarios were used intellectual transposition”. For the “sea battle” scenario, a set of agents
agents that implement various strategies. Using the “sea was developed that implements a number of algorithms
battle” scenario as an example, consider the results of three aimed at solving the problem, two experiments were
different strategies: conducted using different numbers of agents, and the
A. Algorithm based on random events; results of these algorithms on all systems are presented.
B. Algorithm with analyzing field; The developed system was used to conduct an artificial
C. An algorithm that bases its choice on the experience intelligence competition as part of the Bryansk Regional
of previous parties. IT Festival.
As a tournament system, it was circular, i.e. each agent
will play with each other. Acknowledgments
Table 1. Performance characteristics of various algorithms for
The reported study was funded by RFBR, project
the scenario "Sea battle"
number 19-07-00844.
Victories, %
A B C
Defeats, %
References
A 71,7% 93,4%
B 28,3% 70,8% [1] Chesani F., Galassi A., Mello P., Trisolini G. (2017)
C 6,6% 30,2% A Game-Based Competition as Instrument for
Teaching Artificial Intelligence. In: Esposito F.,
As follows from the table, algorithm B and C have a Basili R., Ferilli S., Lisi F. (eds) AI*IA 2017
significant advantage over algorithm A, and algorithm B Advances in Artificial Intelligence. – AI*IA 2017.
is significantly inferior to algorithm C. Lecture Notes in Computer Science, vol 10640. –
Consider a similar table in which copies of these agents Springer, Cham.
participate. [2] Zakharova, A.A., Vekhter, E.V., Shklyar, A.V. (2019)
Table 2. Performance characteristics of various algorithms for The of visualization tools in the meta-design of an
the scenario "Sea battle" with copies of agents educational environment. European Journal of
Victories, % Contemporary Education, Vol. 8 no. 1, pp. 43-51.
A1 А2 B1 В2 C1 С2
Defeats, %
[3] Russian AI Cup – artificial intelligence programming
A1 50,1% 70,9% 72,1% 90,2% 93,1%
competition. – Access mode: http://russianaicup.ru
А2 49,9% 71,3% 70,8% 92,7% 91,9%
В1 29,1% 28,7% 49,5% 69,5% 70,9% [4] Mini AI Cup – artificial intelligence programming
В2 27,9% 29,2% 50,5% 71,1% 70,5% competition. – Access mode: https://aicups.ru/.
С1 9,8% 6,3% 30,5% 28,9% 49,8% [5] Google AI Challenge. – Access mode:
С2 6,9% 8,1% 29,1% 29,5% 50,2% http://ants.aichallenge.org.
[6] Zakharova A.A., Korostelyov D.A., Fedonin O.N.
Based on the second table, we can conclude that (2019) Visualization Algorithms for Multi-criteria
algorithms of the same type on average always have a 50% Alternatives Filtering. Scientific Visualization, Vol.
probability of winning against each other, and the 11, no. 4, pp. 66-80.
conclusion based on the results of the first table is fully [7] Jones, M. T. Artificial Intelligence Programming in
consistent with the conclusion of the second. Applications / M. T. Jones. – M.: infinity science
press llc, 2018. – 312 p.
5. Conclusion [8] Norwing, P. Artificial Intelligence: A Modern
Approach. – UK: Glivice, 2014. – 1408 p.
The developed system allows you to create scenarios
[9] M.B. Mihailuk, P.U. Timohin. (2019) Memory-
with different conditions, as well as set them your own
effective methods and algorithms of shader
unique display and conduct an unlimited number of
visualization of digital core material model. Scientific
tournaments according to the created scenarios. It is also
Visualization, Vol. 11, no. 4, pp. 1-11.
possible to use it as a training system due to special task
scenarios in which the user needs to implement the About the authors
algorithm and send the result to the server for verification.
The article analyzes the main analogues - Russian AI Alena A. Zakharova, Sc.D. in Technique, professor of
Informatics and Software Engineering Department at Bryansk
Cup, Mini AI Cup and Google AI Challenge, identifies the
State Technical University. E-mail: zaa@tu-bryansk.ru
main shortcomings and formulates the properties that the Nikita S. Silchenko, graduate student of Informatics and
platform should have, and provides the platform Software Engineering Department at Bryansk State Technical
architecture that meets all the requirements. University. E-mail: silchenko.nk@gmail.com
Distinctive features of the developed system compared Rostislav A. Krylov, graduate student of Informatics and
to other platforms is its versatility and immutability both Software Engineering Department at Bryansk State Technical
in supported scenarios and in visualization methods for the University. E-mail: oktopy@gmail.com
interaction of intellectual agents. Also, the presence of a Vladimir I. Averchenkov, Sc.D. in Technique, professor of
flexible system for visual reproduction of scenarios allows Computer Technologies and Systems Department at Bryansk
State Technical University. E-mail: aver@tu-bryansk.ru
more efficient development and debugging of algorithms
from various areas of artificial intelligence.