AI-Fuzzy Markup Language With Computational Intelligence For High-School Student Learning

1
AI-FML with Computational Intelligence for High-School Student Learning @ 2021 IEEE CIS Summer School
(The content of this article was also submitted to IEEE CIM on Nov. 5, 2021.)
AI-Fuzzy Markup Language with Computational
Intelligence for High-School Student Learning
Chang-Shing Lee, Senior Member, IEEE, Mei-Hui Wang, Yusuke Nojima, Senior Member, IEEE
Marek Reformat, Senior Member, IEEE, Leo Guo
Abstract—Computational Intelligence (CI), which includes
fuzzy logic (FL), neural network (NN), and evolutionary
computation (EC), is an imperative branch of artificial intelligence
(AI). As a core technology of AI, it plays a vital role in developing
intelligent systems, such as games and game engines, neural-based
systems including a variety of deep network models, evolutionary-
based optimization methods, and advanced cognitive techniques.
The 2021 IEEE CIS Summer School on CI for High-School
Student Learning was held physically at the JanFuSun Resort
Hotel, Taiwan, and virtually on Zoom, on August 10-12, 2021. The
main contents of the Summer School were lectures focused on the
basics of FL, NN, and EC and the workshop on AIoT (Artificial
Intelligence of Things). Invited speakers gave nine courses
covering topics like CI real-world applications, fundamentals of
FL, and the introduction to NN and EC. The 2021 Summer School
was supported by the 2021 IEEE CIS High School Outreach
Subcommittee. We also invited students and teachers of high and
elementary schools from Taiwan, Japan, and Indonesia. They
attended the school and participated in AIoT workshop, gaining
experience in applications of AIoT-FML learning tools. According
to the short report and feedback from the involved students and
teachers, we find out that most participants have quickly
understood the principles of CI, FL, NN, and EC. In addition, one
of the teachers sent the following remark to the organizers: “This
is a great event to introduce students to computational intelligence
at a young age, stimulate them to be involved in rapidly evolving
fields, and foster participation in future research adventures.”
Index Terms—Computational Intelligence, AIoT, AI-FML,
Human and Robot Co-Learning
I. INTRODUCTION
YBRID learning has become a viral approach for
explaining and studying cybernetics systems,
especially when the COVID-19 pandemic has
swamped the world since 2020. This topic has always been
considered as the research domain in the Computational
Intelligence (CI) area. CI is traditionally composed of three
fields: Neural Networks, Fuzzy Systems, and Evolutionary
Computation. CI plays a significant role in developing
successful, intelligent systems - from rule-based systems via
games to cognitive techniques. Some of the most successful AI
systems are based on CI [1].
This work was supported in part by the Taiwan Ministry of Science and
Technology under Grant MOST 110-2622-E-024-003-CC1.
Chang-Shing Lee and Mei-Hui Wang are with the Computer Science and
Information Engineering Department, National University of Tainan, Taiwan
(e-mail: leecs@mail.nutn.edu.tw and mh.alice.wang@gmail.com).
In 2018 and 2019, the IEEE CIS Summer School on CI for
Human and Robot Co-learning was held in Taiwan. However,
owing to the COVID-19 pandemic, the 2020 IEEE CIS Summer
School on CI for Human and Robot Co-learning was moved to
a virtual form using Zoom in Japan and Taiwan. Additionally,
the 2020 CI High School Education Program on CI for AI-FML
Robotic Learning supported by IEEE CIS was held in Taiwan
as a physical tutorial. Both events aimed at promoting the CI
concepts and knowledge to junior high school students and
elementary school students in Japan and Taiwan.
The IEEE CIS High School Outreach Subcommittee’s goal is
to facilitate the outreach to high-school students between the
ages of 12-18 and their teachers. In 2021, IEEE CIS High
School Outreach Subcommittee supported the High School
Engagement @ FUZZ-IEEE 2021 [2] and the 2021 IEEE CIS
Summer School on CI for High-School Student Learning [3].
These events were held physically at the JanFuSun Resort
Hotel, Taiwan, and virtually on Zoom on August 10-12, 2021
[3]. The goal of the summer school was to gather more high-
school students to teach them about the CI knowledge and tools
for real-world applications. Figs. 1(a) and 1(b) show the High
School Engagement @ FUZZ-IEEE 2021 and opening
ceremony @ 2021 IEEE CIS Summer School, respectively.
(a) (b)
Fig. 1. (a) High School Engagement @ FUZZ-IEEE 2021; (b) Opening
ceremony @ 2021 IEEE CIS Summer School.
II. AI-FML WITH IEEE 1855 CO-LEARNING MODEL
A. AI-Fuzzy Markup Language (FML) and Co-Learning
Model
Combining AI with human body language can enhance the
genuineness and understanding of human-robot interactions.
The robot is not just a tool or device. It becomes a reliable
assistant and a friend with whom users can share their thoughts
Yusuke Nojima is with the Computer Science and Intelligent Systems
Department, Osaka Prefecture University, Japan (e-mail: nojima@cs.osakafu-
u.ac.jp).
Marek Reformat is with the Electrical and Computer Engineering
Department, University of Alberta, Canada (e-mail: reformat@ualberta.ca).
Leo Guo is with the NUWA Robotics, Taiwan (e-mail:
guo.leo@nuwarobotics.com).
H

2
and feelings [4].
A Machine-Human Co-learning Model built based on the AI-
FML based on IEEE 1855 is illustrated in Fig. 2. The model is
suitable for learning different age-group students and is
described as follows. The starting point is Human Intelligence
(HI). Then, it unfolds into CI machine learning models,
including perception abilities based on Neural Networks (NN),
cognition skills with Fuzzy Logic (FL), and evaluation methods
based on Evolutionary Computation (EC) techniques. During
the learning process, the model identifies four stages for young
students to gain familiarity with AI-FML: experience-based
learning, operation-based learning, practice-based learning, and
application-based learning. Finally, the students learn with the
support of machines/robots – co-learning – following the
integration of elements of CI with real-life applications and
learning how to interact with the machines/robots.
Fig. 2. Machine-human co-learning model based on AI-FML.
B. Learning Framework of AIoT Workshop based on AI-FML
A learning framework of the AI-FML AIoT (Artificial
Intelligence of Things) workshop focused on the elementary-
school and high-school students is shown in Fig. 3. It can be
described in the following way:
Fig. 3. Learning framework of AIoT workshop based on AI-FML.
1) Students utilize the AI-FML learning platform to construct
the knowledge base and rule base of the CI-based real-world
applications. Additionally, they can apply the Particle Swarm
Optimization (PSO) tool to train the knowledge base (KB) and
rule base (RB) using the training data set. 2) Students make use
of the AI-FML Lab to program the AI-FML Blockly. They can
operate the fuzzy inference mechanism according to the input
data and the learned Adaptive Network-Based Fuzzy Inference
System (ANFIS) model, or PSO-based Fuzzy Markup Language
(PFML)-based KB and RB [5, 6]. 3) Students adopt the NUWA
Lab to interact with the AI-FML robot Kebbi Air. They can co-
learn with the robot to simulate the features of the robot’s
perception and cognition. 4) The inferred results are sent to the
AI-FML robot Kebbi Air through the Message Queuing
Telemetry Transport (MQTT) server. Besides, the AI-FML Lab
and AI-FML learning platform can directly send inferred results
to the AI-FML robot Kebbi Air or the AIoT-FML MoonCar
through the MQTT server and to the AIoT-FML-LT through
HTTP protocol. 5) Finally, the end-user can communicate with
the remote end-user online with such a framework.
III. HIGH SCHOOL STUDENT LEARNING ON CI
A. Information of Participants
The total number of people who registered for the 2021 IEEE
CIS Summer School for High School Student Learning was 214,
including 82, 15, 107, and 10 attendees from Taiwan, Japan,
Indonesia, and other countries, respectively. There were 13
graduates, 93 university students, 31 senior high school
students, 16 junior high school students, 21 elementary school
students, and seven non-students. Figs. 4(a) and 4(b) show the
number of participants from different countries and the number
of participants from other groups of students, respectively. The
percentage of high-school students was around 22%.
(a) (b)
Fig. 4. (a) Number of participants from different countries; (b) Number of
participants from different groups of students.
B. Summer School Program for High School Students
Fig. 5 shows the program for the high school students to learn
the basics of CI at the Summer School. The main contents of
the Summer School are the fundamental pillars of fuzzy
systems, neural networks, and evolutionary computation. The
lectures were designed to be suitable for beginners who wish to
find out more about these areas. The recorded videos of the talks
are available as a set of videos, item No. 1 in the Appendix. The
lectures introduced the basic concepts of CI for young students
or beginners, containing a variety of explanations: (1) capability
of fuzzy logic to compute the degree of human perception of,
for example, such concepts as hot or cold, especially in the case
when different people have different opinions of hot or cold
even at the same temperature; (2) the usefulness of neural
networks as one of the essential models for machine learning
that can compute the mathematical functions; and (3)
explanation of evolutionary computation that is based on the
Learning
Application
Integration
Artificial Intelligence (AI)
Evolution
Cognition
Perception
Neural Network (NN)
AI-FML
IEEE 1855
Elementary School / Experience
Junior High School / Operation
Senior High School / Practice
University / Application
Fuzzy Logic (FL)
Computational Intelligence (CI)
Evolutionary Computation (EC)
Human Intelligence (HI)
Learning Stages
NUWA Lab. @ Japan
Perception, Evolution, and Cognition Perception and Cognition
AI-FML Learning Platform @ Taiwan
AI-FML
Learning Platform
(Taiwan)
NUWA Lab.
(Japan)
Perception, Evolution, and Cognition
AI-FML Lab. @ Taiwan
AIoT-FML-LT
AI-FML Kebbi Air
AIoT-FML MoonCar
AI-FML Lab.
ANFIS/PFML
(Taiwan)
Taiwan
38.3%
Japan
7.0%
Indonesia
50.0%
India
0.9%
Italy
1.4%
Thailand
0.9%
Saudi Arabia
0.5%
Turkey
0.5%
USA
0.5%
Professor/Teacher
15.4%
Graduate
6.1%
University
Student
43.5%
Senior High School
Student
14.5%
Junior High School
Student
7.5%
Elementary School
Student
9.8%
KWS/NUTN Staff
1.9%
Student Parents
1.4%

3
observation of the animals’ behavior patterns and it is one of
the vital machine learning models, too. Integrating emotional
intelligence into existing AIoT applications is yet another
important aspect of AI [5]. In addition to tutorials on basic
concepts of CI, this Summer School provided some lectures on
CI-based real-world applications, as well as a QA system with
linguistic terms and summarization and a workshop on AIoT to
attract high-school students’ attention to learning CI.
Fig. 5. Program for high school students learning CI at the Summer School.
C. Workshop on AIoT
We have designed and developed the AI-FML Lab, NUWA
Lab, and AI-FML learning platform to imitate different aspects
of human behavior. In the three AIoT workshops of the 2021
IEEE CIS Summer School for High-School Student Learning,
we focused on the applications to AIoT-FML intelligent air
conditioner (details are provided in the videos, item No. 2 in the
Appendix), intelligent speaking English, and BCI application,
all based on a fuzzy system. The participants were divided into
small groups using breakout room at Zoom and each group had
a tutor to advise the participants. Fig. 6 shows the AIoT devices
including AI-FML Kebbi Air, AIoT-FML-LT, and AIoT-FML-
MoonCar that are compatible with IEEE 1855-2016 Standard
[7]. Fig. 7(a) shows the AIoT devices at the venue of the 2021
Summer School. The screen image, the top left of Fig. 7(b),
shows a girl tutoring the participants on how to interact with the
AI-FML Kebbi Air using the AI-FML learning platform during
the AIoT workshop.
(a) (b)
Fig. 6. (a) IEEE 1855 @ Kebbi Air and AIoT-FML-LT V3.0. (b) AIoT-FML-
MoonCar.
There were eight high school student teams and two
elementary school student teams. Additionally, when dividing
students into groups, we ensured a balanced diversity of
students in the context of their origin/country. Each group was
asked to perform the following tasks:
- examine the effect of various parameters of fuzzy systems
(e.g., the rule base, the number of membership functions)
and PSO (e.g., the number of particles and the number of
fitness evaluations) on the learning performance.
- design the robot’s expressions according to the
outputs/results of the fuzzy system.
- present or make a short report on what they learned at the
end of the workshop.
The recordings of the participants’ co-learning interaction
with the machines at the AIoT workshop are available, item No.
3 in the Appendix.
(a) (b)
Fig. 7. (a) AIoT devices; (b) Interaction between the participants and the AI-
FML Kebbi Air.
IV. CONCLUSION AND DISCUSSIONS
The 2021 IEEE CIS Summer School for High-School Student
Learning has a significant impact on teaching elements of
Computational Intelligence to high-school or university
students of computer science, mathematics, electrical
engineering, robotics, and related areas. This activity is
supported by CIS, which co-funds the Summer School. The
scientific goal of the School is to promote CI and expand the
attendee’s base from college and graduate students to high-
school or even elementary school students. Further, such an
expansion is in line with the policy of Taiwan and Japan to
promote new courses that include concepts of computational
thinking as a part of the fundamental education in Taiwan and
Japan. Furthermore, it means the involvement of the national
and internationally leading researchers in CI, members, and
senior members of the IEEE. Therefore, we can say that The
Summer School on Computational Intelligence for High-School
Student Learning is a noteworthy example of teaching the
basics of computational intelligence to students from
elementary schools, high schools, and universities in Taiwan
and Japan. Further, we can say that the AIoT Workshop gives
students a good experience in applying the AI-FML Robot and
CI-based AIoT-FML Learning Tool for real-world
applications.
The total number of individuals who attended at least two-
thirds of lectures and the workshops on AIoT was 78, including
48, 9, and 21 from Taiwan, Japan, and Indonesia. All high-
school students received the certificate. The ratios of students
awarded a certificate of participation was 0.585, 0.600, and
0.196 from Taiwan, Japan, and Indonesia, respectively. The

4
total number of attendees who submitted a short report and
feedback survey was 80 and 77, respectively. The submitted
reports and surveys showed that participants understood
principles of computational intelligence, neural networks, fuzzy
logic, and evolutionary computation to a higher degree than in
the case of previously organized schools. The targeted audience
of this Summer School was high school students. Yet, the
registered individuals represented various people from teachers,
via graduate students to elementary school students. From our
experience and feedback, categorization (e.g., target age)
according to the background knowledge is very important for
both lecturers and participants to make summer schools fruitful.
We think this event was a well-organized Summer School.
We were happy to see that the School was attended by many
participants, both students and teachers, from Taiwan, Japan,
Indonesia, and Europe. Finally, we would like to summarize the
report citing some feedback comments we received: “This is a
great event to introduce students to computational intelligence
at a young age, stimulate them to be involved in rapidly
evolving fields, and foster participation in future research
adventures.” And further, “it is in line with IEEE’s mission to
advance technology for humanity. Moreover, IEEE offers
STEM education opportunities for pre-university students &
teachers across the globe. After joining this event, hope they
have a big passion & feel motivated to be involved in global
issues & trends.”
ACKNOWLEDGMENT
The authors would like to thank for the support provided by
the IEEE CIS High School Outreach Subcommittee, the
Ministry of Science and Technology (MOST) in Taiwan,
JanFuSun Resort Hotel, E. Sun Commercial Bank, Zsystem
Technology Company, NUWA Robotics, Tainan City
Government, and Kaohsiung City Government in Taiwan.
APPENDIX
TABLE I. SHORT TEXTUAL DESCRIPTIONS OF THE VIDEOS.
No Descriptions
1
Topic: Slide and Recorded Videos
https://sites.google.com/asap.nutn.edu.tw/2021-ieee-cis-summer-
school/lecture-materials
Topic: Onsite Recorded Videos
school/videos/recorded-videos
Topic: Fast-Forward Videos
school/videos/fast-forward-videos
2
Topic: AIoT-FML intelligent air conditioner application
demonstration
Link: https://youtu.be/MU_YjneyDbQ
Descriptions: Chia-Ying Liu demonstrates AIoT-FML intelligent
air conditioner application.
3
Topic: 2021 IEEE CIS Summer School Workshops on AIoT
Link: https://youtu.be/Q0-6FYt1394
Descriptions: This video shows the participants’ co-learning
situation with the machines at the AIoT workshop.
Note: This YouTube video provides English subtitles.
REFERENCES
[1] IEEE CIS, “What is CI?” Oct. 2021. [Online] Available:
https://cis.ieee.org/about/what-is-ci.
[2] FUZZ-IEEE 2021, “High-School Engagement, Jul. 2021. [Online]
Available: https://attend.ieee.org/fuzzieee-2021/high-school-
engagement/.
[3] OASE Lab., NUTN, “2021 IEEE CIS Summer School on Computational
Intelligence on High-School Student Learning,” Aug. 2021. [Online]
Available: https://sites.google.com/asap.nutn.edu.tw/2021-ieee-cis-
summer-school.
[4] NUWA Robotics, “Kebbi Air,” Aug. 2021. [Online] Available:
https://www.nuwarobotics.com/en/.
[5] C. S. Lee, M. H. Wang, W. K. Kuan, S. H. Huang, Y. L. Tsai, Z. H. Ciou,
C. K. Yang, and N. Kubota, “BCI-based hit-loop agent for human and AI
robot co-learning with AIoT application,” Journal of Ambient Intelligence
and Humanized Computing, 2021. (DOI: 10.1007/s12652-021-03487-0)
[6] C. S. Lee, Y. L. Tsai, M. H. Wang, S. H. Huang, M. Reformat, and N.
Kubota, “Adaptive fuzzy neural agent for human and machine co-
learning,” International Journal of Fuzzy Systems, 2021. (Accepted in
Sept. 2021)
[7] G. Acampora, B. di Stefano, and A. Vitiello, “IEEE 1855TM: The first
IEEE standard sponsored by IEEE Computational Intelligence Society,”
IEEE Computational Intelligence Magazine, vol. 11, no. 4, pp. 4-7, 2016.

AI-Fuzzy Markup Language With Computational Intelligence For High-School Student Learning

More Related Content

Similar to AI-Fuzzy Markup Language With Computational Intelligence For High-School Student Learning (20)

More from Wendy Berg (20)

Recently uploaded (20)

AI-Fuzzy Markup Language With Computational Intelligence For High-School Student Learning