Conference PaperPDF Available

An adaptive paradigm for smart education systems in smart cities using the internet of behaviour (IoB) and explainable artificial intelligence (XAI)

Authors:
1
An adaptive paradigm for smart education
systems in smart cities using the internet of
behaviour (IoB) and explainable artificial
intelligence (XAI)
Dr. Ossama Embarak
Higher Colleges of Technology
Fujairah, UAE
ORCID: 0000-0003-4504-8561
oembarak@hct.ac.ae
AbstractThe rapid shift towards smart cities, particularly in
the era of pandemics, necessitates the employment of e-learning,
remote learning systems, and hybrid models. Building adaptive
and personalized education becomes a requirement to mitigate the
downsides of distant learning while maintaining high levels of
achievement. Explainable artificial intelligence (XAI), machine
learning (ML), and the internet of behaviour (IoB) are just a few
of the technologies that are helping to shape the future of smart
education in the age of smart cities through Customization and
personalization. This study presents a paradigm for smart
education based on the integration of XAI and IoB technologies.
The research uses data acquired on students' behaviours to
determine whether or not the current education systems respond
appropriately to learners' requirements. Despite the existence of
sophisticated education systems, they have not yet reached the
degree of development that allows them to be tailored to learners'
cognitive needs and support them in the absence of face-to-face
instruction. The study collected data on 41 learner's behaviours in
response to academic activities and assessed whether the running
systems were able to capture such behaviours and respond
appropriately or not; the study used evaluation methods that
demonstrated that there is a change in students' academic
progression concerning monitoring using IoT/IoB to enable a
relative response to support their progression.
Keywords: Smart education, Explainable artificial intelligence
(XAI), the internet of behaviour (IoB) , Smart cities education,
adaptive learning, future AI-based education.
1. Introduction
Education and learning are no longer confined to the confines
of schools and classrooms. Technology's advancement has
altered the way educational institutions operate. The Internet
of Things (IoT) has the ability to connect a wide range of
objects to the Internet. In a wireless context, artificial
intelligence (AI) and machine learning (ML) have the ability
to make computers think like and mimic people. These
technologies have an impact on every aspect of people's life.
They have a wide range of applications in education,
manufacturing, healthcare, transportation, smart cities, and
energy.
The education/learning area is not immune to technological
advancement. The Internet of Things has played a significant
role in engaging learners in the classroom. The Internet of
Things has assisted instructors in making their classrooms
more dynamic and interesting. IoT makes education more
accessible in terms of ability, geography, and socioeconomic
condition. Immersion, for example, is a strategy used to learn
a foreign language based on real-time input. Learning a
language in its home country is simply because the speaker's
feedback is openly available. It is difficult to construct similar
environments outside of the country, but IoT aids in creating
simulation environments in which students can be watched
and teachers can provide real-time feedback to pupils. IoT
networking can facilitate task-based learning. Students learn-
by-doing in doing task-based learning, which allows teachers
to support and monitor a student's performance
automatically. IoT has been a big help to impaired pupils by
giving technical assistance.
Along with IoT, AI and ML have made significant
contributions to advancements in research and education. AI
and ML aid in the automation of time-consuming class
operations such as attendance and grading. Educational
Software can be customized to meet the demands of the
student. IoT and AI aid with e-learning courses by allowing
students to access lessons and videos from any location and
at any time. Discussion groups and forums also allow
students to discuss questions, ideas, and information.
The role of education in propagating knowledge has grown
in importance over the last several years due to the
fulminatory proliferation of knowledge. Meanwhile, the
educational process's paradigm is transforming, requiring
different pupils to complete their learning in a variety of
ways. As a result, a smart educational environment is
promoted. It integrates a variety of information and
communication technologies to energize the learning process
and adapt to the unique needs of individual pupils. The
quality of students' learning processes can be improved by
continuously monitoring and evaluating their states and
actions via information sensing devices and information
processing platforms in order to provide feedback on their
learning processes. The Internet of Things is committed to
achieving significant variation in life, individual well-being,
and organizational productivity. The IoT has the potential to
enable expansions and enhancements to critical utilities in a
variety of industries by creating a unique environment for
application development. Applying the Internet of Things
concept to any educational environment would improve the
quality of education by allowing students to learn more
quickly and teachers to do their duties more efficiently. It is
delicate to attempt to influence and modify user behaviour,
2
since it may face resistance and other psychological issues
connected to comfort and trust. Emerging technologies such
as XAI will aid in the user's comprehension and trust of any
system that makes use of AI models. XAI's goal is to employ
methods and approaches to communicate to the user what an
AI model does and why resulting in a complete understanding
of system operations. As a result, tracking, analyzing, and
influencing user behavior will become significantly simpler
[1] .
However, very few studies focus on how to leverage the
internet of behaviours (IoB) to personalize learning content
and how IoB can be used to monitor students' behaviour that
affects their attainment, progression, and performance. This
work will demonstrate a proposed paradigm for integrating
explainable artificial intelligence (XAI) and the internet of
behaviour (IoB) to tailor learning content to students'
cognitive abilities and automate academic progress tracking.
This article aims to employ a combination of the IoB and XAI
to create trustworthy and intelligible frameworks that operate
in the education domain where users' behaviour changes. It
offers an IoB-based XAI-based system for use in the
education sector to change users' behaviour toward an eco-
friendly one to maximize the learning attenuation and robust
student’s academic progression. This framework
incorporates IoT, AI, Data Analytics, Behavioral Science,
and XAI approaches to deliver smart, adaptable education.
The following summarizes the paper's contributions:
Propose a trustworthy and understandable IoB-XAI-
based education system capable of influencing and
changing learners' habits and behaviour for their
educational benefit.
Provide a scenario illustrating the learner's
behaviour inside the class, practicals, and other
related activities to create knowledge about the
importance of controlling a student's progression.
Compare and contrast the primary characteristics of
existing systems and the proposed paradigm.
Outline prospective advancements and the planned
system's future directions.
Even though various studies seek to explain how educational
improvements occur, this study examines how explainable
artificial intelligence and the internet of behaviours can be
used to tailor and modify learning resources to students'
cognitive abilities and requirements. This paper is organized
as follows. This paper begins with the abstract and
summarizes the main concern and discoveries of the work.
Second, the introduction summarises the major research gaps
and our main contributions and aims. Thirdly, the literature
review where the context in which the associated work is
presented and elaborated. Fourthly, the research employed a
methodology that includes a detailed description of the
research concept, participants, data collection techniques,
and materials used in the study. Suggestions and
contributions are made in light of the findings. Finally, the
expansion that will be performed in the future.
2. Literature review
The Internet of Things (IoT) is a concept that allows diverse
devices to connect and interact with each other. This can
enhance education, healthcare, transportation, and other
commercial operations [2]. They might be virtual
(information world) or physical (physical world). Objects can
self-identify and integrate with the communication layer.
From a psychological standpoint, the Internet of Behavior is
a subset of the Internet of Things (IoT) that tries to address
users' behaviours [3]. There is a growing body of research on
the Internet of Behavior (IoB) and how it might help
businesses understand their customers better [4]. The IoB
platform helps academic institutions fully comprehend their
pupils. By connecting all phones in the app, users can see
their swing and stroke problems and get visual suggestions
on how to fix them. The connection of gadgets creates new
data points that go beyond the IoT. (IoT). Client data is shared
across linked devices in real-time.
The COVID-19 pandemic has influenced various aspects of
consumer behavior, including brand loyalty, staff
productivity, and customer engagement. The environment,
technology, and human health are all affected. As a result,
tracking people's behavior is crucial in bad situations. Using
machine learning for activities like mask recognition might
help enforce laws and detect negligence [5]. There will be
chances to profit from the information obtained by assessing
the history of patterns in various companies, societal, health,
political, and other realms. A person's behavior might impact
their willingness to cooperate or collaborate. Regardless of
the other four factors: cognition, emotion, personality [6], and
interpersonal communication, behavior is responsible for the
tendency to act. In this way, focusing on behavior might help
us understand how to impact and treat the individual [7]. The
Multi-layered paradigm promotes generating learning
materials based on learners' personalities and learning
patterns [8]. Phone sensors can now assess biometrics and
healthy behavior, allowing businesses to develop
applications that care about users' health. For example, health
applications track sleep, heart rate, and respiration. After
analyzing the user's behavior, such applications provide
alerts, messages, and suggestions to improve sleep and
motivate them to achieve daily objectives [9].
During the COVID-19 epidemic, various nations created
health smartphone applications to help residents stop the
virus's spread. For example, Code-based apps track a user's
travel, contact history, and physical biometrics such as
temperature. Then, a colourful QR code is created to identify
his health state. As a result, various constraints may apply to
the user and influence his behaviour, such as permitting travel
or quarantine at home or a central area [10]. Popular
recommendation systems employ user behaviour, notably
viewing history and clicks, to improve an app or website's
experience. Their technology uses deep neural networks to
evaluate data and recommend films and shows to users. Also,
virtually all social media programs provide feeds depending
on viewing habits, time spent on one account, or interest in
subscribing to certain channels [11]. In transportation, IoB
can assist both drivers and passengers. Uber employs
gamification to influence driver behaviour after a lengthy
history of driver conflicts and high turnover. Uber uses loss
aversion, recognition, and intrinsic motivation to penalize
drivers [12]. The Deep Reinforcement Learning algorithms
(Deep RL) were used to tackle the power consumption based
3
on the user's behaviours [13]. Some systems tried to
understand student’s learning styles to generate
recommendations [14]. IoB plays a critical part in
discovering learners' behaviours and therefore eliminating
any misleading ratings that may occur intentionally while
producing recommendations [15].
3. The proposed model
Learners vary In their capabilities and cognitive styles, which
Imposes the need to personalize learning resources to fit their
abilities and needs. The use of IoT is helpful to collect
different forms of data that could be textbase, images, or
audio data. This study focuses on collecting data about
students and teachers and the learning environment in a
university. IoB has just been identified as one of the top
technological trends. The COVID-19 outbreak is mostly
responsible for IoB becoming a trend since it has transformed
how consumers interact with brands, forcing businesses to
reevaluate how they connect with customers.
3.1 The role of IoB.
The IoB idea seeks to correctly evaluate data and apply that
understanding to construct and provide new things from the
perspective of human psychology. The IoB seeks to
comprehend data gathered from online user activities through
the lens of behavioural psychology. It tries to answer
concerns about how to evaluate data and use that knowledge
to design and deliver new services, all based on human
psychology. This new trend may influence Quality
Infrastructure since many firms may improve their
connectivity, resulting in higher user demand. As represented
by pyramid figure 1, the IoT changes data to information, and
the IoB may transform our knowledge into true wisdom.
Fig. 1. IoB and IoT In Smart layered System
The following demonstrates each layer's main function
1. Observe and track learner behavior: The Internet of
Things (IoT) is used to collect information about
students inside the classroom, their learning
activities, their volunteering activities and
behaviours, how they react to the exercises and
formative assessments, class collaboration work,
hands-on activities, and everything else that is
related to their academic progression.
2. Analyze students' behaviour: the data obtained from
IoT may be utilized to analyse the student's
behaviour appropriately. However, this analysis will
not deliver automatic suggestions through the
system but provide facts about learners' actions
throughout their study and provide some
explanation of some phenomena.
3. Recognize significance: Data acquired by the
Internet of Things has been enriched by behavioral
and psychological analysis (IoT). With the Internet
of Behaviour, companies and organizations
worldwide may better understand and anticipate
how certain behaviors lead to certain results. The
IoB concept allows academic institutions to measure
and predict future student performance. Academic
institutions will be able to respond to students'
behavior by making recommendations and tracking
their progress. The IoB will continue to collect data
on students' choices and patterns of behavior. It's
possible to acquire a far better picture of students'
behavior by analyzing this data than previously
possible.
4. Significance influence: the IoB mechanism can be
used by academic institutions to maintain an
adaptive system that recognizes the behaviours of
students and advises behavioural modifications that
will result in a better outcome at this level.
3.2 Explainable AI and IoB for a smart learning
education system
As the fourth industrial revolution begins, we are seeing a
rapid and broad deployment of artificial intelligence (AI),
contributing to a more algorithmic society. Even with these
extraordinary developments, lack of transparency is a major
barrier to AI-based systems' usage. They can make
tremendous predictions, but they can't be explained. This has
rekindled the debate about explainable AI (XAI) [16]. A
study field might significantly improve AI-based system trust
and transparency. It is required for AI to evolve steadily and
unhindered. Explainable AI helps use AI for data mining
while maintaining transparency. To make AI discoveries
more intelligible to humans, researchers have developed
XAI, a new field of study. Researchers have been studying
the subject of explainable for decades when they looked at
explanations for expert systems [17]. A direct result of
AI/relentless ML's widespread effect on critical decision-
making processes, but without providing explicit information
about the reasons behind specific judgements, proposals,
projections or actions. Increasing social, ethical, and legal
pressures necessitate new AI techniques that can explain
decisions.
4
Technically, there is no accepted definition of explainable AI.
Rather than a specific technology concept, XAI refers to a
movement, initiatives, and actions addressing AI
transparency and trust. Thus, XAI is a new breed of
artificially intelligent companions that humans can
comprehend and control while keeping great learning
performance (prediction accuracy) [18].
3.2.1 Data collection: data can be collected from various
related IoT devices such as wearable watches,
mobiles, cameras, campus cameras, labs monitor
devices, temperature sensors, etc. All forms of
students' activities such as volunteering works,
public speeches, seminars, presentations, project
shows/exhibitions etc., must be under spot to
understand students' behaviours and hence used as
input for analysis and predictions using XAI
models.
Fig. 2. IoB via IoT for a smart learning education system
3.2.2 IoB via XAI in a smart education system
The collected data is then used for applying XAI models,
which aims to
1. Automate the mapping of learner behaviour to
particular symptoms based on gathered data.
2. Establish behavioural patterns for tailoring learning
materials to individual requirements.
3. Justify rationales for specific model outcomes.
4. Provide all the necessary information for managing
patterns to avoid any projected faults or
vulnerabilities influencing model performance.
5. Revealing unanticipated patterns, rules, and reasons
for certain occurrences.
6. Create a smarter model based on a better knowledge
of the rationales for certain results.
Fig. 3. IoB objectives in smart learning education
Automate learner’s behaviour mapping: The main purpose
of IoB in education is to map particular behaviour to a
specific intellectual style. Several studies tried to find the
learning differences between learners [19]. The use of IoB
will automate the process to build smarter education
systems.
Establish behavioural patterns: Understanding learners'
behaviours will help maintain patterns of each behaviour
and better understand learners' perceptions and intellectual
styles.
Justify model outcomes: Several debates have erupted in
recent years about AI/ML-enabled systems producing
biased or discriminating outcomes [20]. That means there
will be a greater need for explanations to guarantee that AI-
based judgments are not made incorrectly. We mean the
need for explanations or justifications for that specific
conclusion rather than describing the inner workings or
logic of reasoning underlying the decision-making process.
Using XAI systems gives the necessary knowledge to justify
results, especially when making unexpected judgments. It
also assures an auditable and demonstrable mechanism to
explain algorithmic conclusions as fair and ethical, leading
to the development of trust.
Control the model: It is crucial not just defending decisions
but to provide the required prevention of errors. Indeed,
learning more about system behaviour offers better insight
into undiscovered vulnerabilities and defects and aids in
speeding and correcting errors in low-criticality scenarios
(debugging) for better model controlling.
Reveal discoveries: The suggested automated model, which
is based on student behaviour, would aid in the discovery of
numerous hidden patterns and rules that will lead to
improved performance.
5
Improve smarter model: Maintaining automated
behavioural patterns through XAI will help build a smarter
system that is upgraded regularly. A model that is easy to
describe and understand can be improved more readily.
Users will be able to make the system smarter since they
understand why it generated specific outcomes. Thus,
IoB/XAI serve as the cornerstone for continual human-
machine iteration and development.
Learning technologies such as Nearpods and BBL
collaboration were utilized to monitor the responses of 41
students (students in programming courses) to activities
such as polls and evaluate if the system supports related
activities. The collaborative work and game-
based/formative evaluations where students can answer
using their mobiles and tablets. Students were also asked to
complete a survey on their volunteer work; the following
table compiles the summary. Table 1 shows students'
capabilities and social keens in their academic performance;
we segment students into four groups, Elite (E), Fair(F),
Below-average (B), and Severely weak(S). We mined more
to understand these students' behaviours. The major goal is
to evaluate the system's ability to understand student
behaviour and personalise learning resources to each
student's strengths, needs, skills, and interests.
Table 1: Learner's capabilities and Social keens
symptoms
The following table illustrates several segments of
learners based on their speaking and writing
talents and their collaboration and volunteering
interests.
Table 2: Learner's segmentation
The following equations summarize the equation used to
measure if the system support
Where,
SP(ES) is the system attractiveness for elite learners;
calculated SP(ES) = 16%
SP(FS) is the system attractiveness for fair learners
calculated SP(FS) = 34%
SP(AS) is the system attractiveness for below-average
learners; calculated SP(AS) = 11%
SP(WS) is the system attractiveness for weak learners;
calculated SP(WS) = 39%
4. Discussions
IoT can help academic institutions gather data on student
reactions to events and understand academic
accomplishment. The learning process will help them
overcome their deficiency. In the first round, we evaluate the
system's ability to engage students in activities without their
6
knowledge and if instructor monitoring via technology
influences their academic achievement. SP(ES) = 16%,
SP(FS) = 34%, SP(AS) = 11%, and SP(WS) = 39%. Given
that the major purpose is to utilize IoB to tailor learning
resources and XAI to explain how the system reacts, the
system should help students by categorizing them and
picking evaluative elements. As part of the second cycle, we
taught students about assessment factors: Speaking
(challenging teachers, seldom asking questions), Writing
(Sentence with elaboration, Short sentences habits).
Volunteering, Collaboration (leading the work, quietness)
(Heavenly involved in activities, Focusing on the given task).
Quiet pupils and those who were focused on the activity were
also given recommendations to improve their elaboration
abilities. SP(ES) increased by 26%, SP(FS) by 38%, SP(AS)
by 19%, and SP(WS) by 17%. This shows that students may
conduct wisely in academic activities if they have a tailored,
regulated framework for learning.
5. The challenges and limitations
The study had several challenges due to many factors: time,
involved students, students selections, and measured
features. The following summarizes a few of these
limitations.
- Time constraints, since we collected data from 41
students over one semester.
- Few features were identified, and we intend to add
more in future study extensions.
- Some students were hesitant to participate in the
research after learning about it.
- The study solely includes students from the
computing division.
6. The fututre directions
The IoT is particularly useful for collecting data about
learners; the IoB is utilized to analyze how pupils react in
response to academic activities. We intend to consider more
factors and incorporate students from other divisions. In
addition, we will create an automated software that will allow
students to understand the elements influencing their
performance and tailor their learning materials. Use a smart
education system that limits students' development from one
level to the next based on their abilities and knowledge
leverage, as determined by automated algorithms that respect
their scraps of skills and knowledge.
7. The conclusion
The study tried to figure out how a smart education system
can make learning materials more personalized for each
student. The proposed method uses IoT and IoB to get
information about students and learn about their behaviour.
Also, using XAI to filter/select features that students can
choose to check their performance and get the help they need
from the education systems is a good thing. Students can
grow in the future by adding more features and making the
implementation broader so that more students from different
divisions can use it.
References
[1] D. Gunning, D.A.-A. magazine, undefined 2019, “DARPA’s
explainable artificial intelligence (XAI) program,” Ojs.Aaai.Org, 2019.
[2] M. Abdel-Basset, G. Manogaran, M. Mohamed, E. Rushdy, “Internet
of things in smart education environment: Supportive framework in the
decision-making process,” Concurrency and Computation: Practice
and Experience, 31(10), 2019, doi:10.1002/cpe.4515.
[3] A. Halgekar, A. Chouhan, I. Khetan, J. Bhatia, N. Shah, K. Srivastava,
“Internet of Behavior (IoB): A Survey,” Ieeexplore.Ieee.Org, 16,
2022, doi:10.1109/iscon52037.2021.9702450.
[4] H. Elayan, M. Aloqaily, M. Guizani, “Internet of Behavior (IoB) and
Explainable AI Systems for Influencing IoT Behavior,” Arxiv.Org,
2021.
[5] M. Moghaddam, H. Muccini, … J.D. preprint arXiv, undefined 2022,
“Designing Internet of Behaviors Systems,” Arxiv.Org, Mar. 2022.
[6] O.H. Embarak, “Three Layered Factors Model for Mining Students
Academic Performance,” in 2018 Fifth HCT Information Technology
Trends (ITT), IEEE: 219226, 2018.
[7] X. Zhou, W. Liang, K.I.K. Wang, H. Wang, L.T. Yang, Q. Jin, “Deep-
Learning-Enhanced Human Activity Recognition for Internet of
Healthcare Things,” IEEE Internet of Things Journal, 7(7), 64296438,
2020, doi:10.1109/JIOT.2020.2985082.
[8] O. Embarak, “Multi-Layered machine learning model for mining
learners academic performance,” Advances in Science, Technology
and Engineering Systems, 6(1), 850861, 2021,
doi:10.25046/aj060194.
[9] I. Sadek, A. Demarasse, M. Mokhtari, “Internet of things for sleep
tracking: wearables vs. nonwearables,” Health and Technology, 10(1),
333340, 2020, doi:10.1007/s12553-019-00318-3.
[10] I. Sadek, S.U. Rehman, J. Codjo, B. Abdulrazak, “Privacy and Security
of IoT Based Healthcare Systems: Concerns, Solutions, and
Recommendations,” Lecture Notes in Computer Science (Including
Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in
Bioinformatics), 11862 LNCS, 317, 2019, doi:10.1007/978-3-030-
32785-9_1.
[11] A. Da’u, N. Salim, I. Rabiu, A. Osman, “Recommendation system
exploiting aspect-based opinion mining with deep learning method,”
Information Sciences, 512, 12791292, 2020,
doi:10.1016/j.ins.2019.10.038.
[12] N. Scheiber, “How Uber Uses Psychological Tricks to Push Its Driver,”
New York Times, 1, 2017.
[13] S.H. Kim, C. Lee, C.H. Youn, “An accelerated edge cloud system for
energy data stream processing based on adaptive incremental deep
learning scheme,” IEEE Access, 8, 195341195358, 2020,
doi:10.1109/ACCESS.2020.3033771.
[14] O. Embarak, “A New Paradigm through Machine Learning: A
Learning Maximization Approach for Sustainable Education,”
Procedia Computer Science, 191, 445450, 2021,
doi:10.1016/j.procs.2021.07.055.
[15] O. Embarak, “Demolish falsy ratings in recommendation systems,” in
2019 Sixth HCT Information Technology Trends (ITT), IEEE: 292
295, 2019.
[16] O. Embarak, “Explainable Artificial Intelligence for Services
Exchange in Smart Cities,” Explainable Artificial Intelligence for
Smart Cities, 1330, 2021, doi:10.1201/9781003172772-2.
[17] A. Das, P. Rad, “Opportunities and Challenges in Explainable
Artificial Intelligence (XAI): A Survey,” Arxiv.Org, 2020.
[18] F. Xu, H. Uszkoreit, Y. Du, W. Fan, D. Zhao, J. Zhu, “Explainable AI:
A Brief Survey on History, Research Areas, Approaches and
Challenges,” Lecture Notes in Computer Science (Including Subseries
Lecture Notes in Artificial Intelligence and Lecture Notes in
Bioinformatics), 11839 LNAI, 563574, 2019, doi:10.1007/978-3-
030-32236-6_51.
[19] O. Embarak, K. Saeed, M. Ali, “Perceiving Intellectual Style to Solve
Privacy Problem in Collaborative Systems,” Lecture Notes on Data
Engineering and Communications Technologies, 29, 401410, 2019,
doi:10.1007/978-3-030-12839-5_37.
[20] R. Confalonieri, L. Coba, B. Wagner, T.R. Besold, “A historical
perspective of explainable Artificial Intelligence,” Wiley
Interdisciplinary Reviews: Data Mining and Knowledge Discovery,
11(1), 2021, doi:10.1002/widm.1391.
... While education systems have progressed, personalization to meet students' cognitive needs during non-face-to-face instruction remains a frontier. Various technologies are revolutionizing education in smart cities, enabling personalized learning and customized content based on individual preferences [42,43]. Leveraging technologies such as the Internet of Things in educational settings not only accelerates students' learning but also significantly enhances instructors' effectiveness (3,9066). ...
... Utilizing advanced technologies, smart education streamlines the processes of teaching, learning, communication, and collaboration, leading to increased efficiency due to timely notifications [44]. [42,43] proposes a transformative paradigm for smart education based on the integration of XAI (Explainable Artificial Intelligence) and IoB (Internet of Behavior) technologies (IoT and IoB) to collect and analyze student behavior data. XAI further refines aspects for students to monitor their performance, ensuring tailored aid from the educational system [43]. ...
... OLMs could be a form of XAI for an ITS but suited to user needs. One interesting work regarding adaptive to user needs XAI can be considered the work by Embarak [34]. In this work, an intelligent education system was examined. ...
Conference Paper
Full-text available
Nowadays, the need to explain the decisions or predictions made by Artificial Intelligence (AI) is emerging more than ever as AI applications are more complex. The research field of eXplainable Artificial Intelligence (XAI) tries to fulfill this need. XAI provides a way to help humans understand how an AI’s predictions and decisions come. The scope of this work is to examine the role of XAI in the field of Education, especially in Educational Data Mining in Vocational Education and Training.
... Additionally, the study explored whether or not operating systems were able to accurately capture and respond to academic behaviours. The research employed evaluation methods that established a direct connection between the adoption of Internet of Things and Internet of Business monitoring and the academic improvement of students, hence providing a proactive support system for the success of the students (Embarak, 2022). ...
Article
Full-text available
AI, or artificial intelligence, refers to the process of using computers and other technologies to make them smarter than people and to assist them in solving problems that they encounter on a daily basis. Some examples of artificial intelligence technologies are machine learning and deep learning. These technologies make use of algorithms to improve their ability to predict future events without the assistance of a human being. Explainable Artificial Intelligence, often known as XAI, is a sort of artificial intelligence that is capable of explaining to humans the reasoning behind a decision or a prediction that it has made. When it comes to crucial duties like security, healthcare, or money, the objective of XAI is to make artificial intelligence systems more open, trustworthy, and accountable. This is especially important when these systems are employed. Within the scope of this essay, an orderly evaluation of the literature on XAI methodologies that can be applied in a variety of contexts, particularly within the education sector, is presented. The purpose of this study is to initiate a comprehensive investigation of XAI in the field of education, with the goal of examining its significance, potential applications. This will be accomplished by examining the current situation as well as possible future advances of XAI in education. We believe that our systematic review contributes to the existing body of research on XAI by pointing the way for additional research to be conducted in this field.
... It becomes necessary to develop adaptive and personalized education to lessen the drawbacks of distance learning and keep high standards of accomplishment. ML, the Internet of behavior, and explainable AI are just a few of the technologies influencing the direction of smart education in the era of EL and MOOC through personalization and customization [11]. ...
Article
Full-text available
Smart strategies and intelligent technologies are enabling the designing of a smart learning environment that successfully supports the development of personalized learning and adaptive learning. This trend towards integration is in line with the growing prevalence of Internet of Things (IoT)-enabled smart education systems, which can leverage Machine Learning (ML) techniques to provide Personalized Course Recommendations (PCR) to students. Furthermore, the existing recommendation techniques are based on either explicit or implicit feedback and fail to capture the changes in learners’ preferences while integrating implicit or explicit feedback. To this end, this paper proposes a new model for personalized learning and PCR that is enabled by a smart E-Learning (EL) platform. The model aims to gather data on students’ academic performance, interests, and learning preferences and utilize this data to recommend the courses that will be most beneficial to each student. The proposed approach makes suggestions based on the learner’s interactions with the system and the cosine similarity in related contents by combining explicit (user ratings) and implicit (views and behavior) methodologies. The suggested method makes use of ML algorithms and an EL Recommender System (RecSys) based on Collaborative Filtering (CF).This includes Random Forest Regressor (RFR), Decision Tree Regressor (DTR), K-Nearest Neighbors (KNN), Singular Value Decomposition (SVD), eXtreme Gradient Boosting Regressor (XGBR), and Linear Regression (LR). The proposed solution is benchmarked against existing approaches on both predictive accuracy and running time. Experimental results are conducted based on two benchmark datasets (Coursera and Udemy). The proposed model outperforms existing top-K recommendations techniques in terms of accuracy metrics such as precision@k, Mean Average Precision (MAP)@k, recall@k, Normalized Discounted Cumulative Gain (NDCG)@k, Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) for PCR. From the experiments, it can be shown that SVD can perform well in terms of higher accuracy and MAP and NDCG and lower MAE, RMSE, and MSE values when contrasted to other proposed algorithms because it is better suited to capture complex student-course interactions. The proposed solutions are promising on two different datasets and can be applied to various RecSys domains.
Chapter
IoB signifies a revolutionary growth in the digital age, expanding the potential of IoT by capturing & evaluating human behaviors through digital interactions. ML can offer novel insights that can improve patient outcomes & increase the effectiveness of IoB-based medical & healthcare procedures, which in turn can help with diagnosis, treatment protocols, & clinical decision-making. Emergence of the IoB has created new avenues for influencing and assisting human decision-making. IoB describes how digital data produced by human activity is gathered, analyzed, & used to support decision-making. Our lives could be profoundly impacted by this technology in a number of areas, including healthcare, education, finance, and transportation. This chapter discuss how IoB integrates with Explainable AI methodologies to offer a reliable experience in the process of modifying IoT behavior, which in turn influences user behavior. The IoB idea, essential elements, architecture, auxiliary technologies, applications, and problems are covered in this chapter.
Article
Full-text available
The emergence of the Internet of Behaviors (IoB) has created new opportunities for influencing and guiding human decision-making. IoB refers to the collection, analysis, and application of data generated by individuals' online activities, behaviors, and interactions. This concept integrates data from various sources, including social media, wearable devices, smartphones, and other digital platforms, to gain insights into human behavior patterns. This technology can profoundly affect various areas of our lives, such as healthcare, education, and transportation. This paper explores the transformative potential of IoB in educational businesses, where it enables personalized learning, real-time feedback, and improved student retention. By analyzing data on student engagement and performance, IoB supports differentiated instruction, enhances collaborative learning, and drives data-driven curriculum development. Additionally, IoB contributes to students’ health and safety through wearable technology and promotes smart, resource-efficient classrooms. However, the implementation of IoB in education poses significant challenges, including privacy concerns, technical complexities, and access disparities. The paper identifies key areas for future research, such as the integration of IoB with traditional pedagogical approaches, equitable access to IoB technologies, and development of ethical standards to safeguard student privacy. This commentary underscores IoB's potential to revolutionize education while emphasizing the need for careful consideration of its challenges to ensure broad and equitable benefits.
Chapter
The increasing use of AI in modern smart cities calls for explainable artificial intelligence (XAI) systems that can improve the efficiency and effectiveness of city operations while being transparent, interpretable, and trustworthy. Developing a unified framework for XAI that can handle the heterogeneity of data and systems in smart cities is the first challenge, considering the need to incorporate human factors and preferences in AI systems. The second challenge is developing new XAI methods that can handle the complexity and scale of smart city data. Addressing ethical and legal aspects is also critical, including ensuring that AI systems are fair and unbiased, protecting citizens' privacy and security, and establishing legal frameworks. Evaluating the effectiveness and usability of XAI systems is also crucial in improving city operations and stakeholder trust apart from XAI research for smart cities: improved visualization, human feedback, integration.
Chapter
AI systems are integral to the development of smart cities, but their complexity can make decision-making processes opaque, leading to concerns about accountability and transparency. Explainable AI (XAI) aims to address this by designing algorithms that can explain decisions in a way that humans can understand. XAI can increase transparency and accountability in smart cities, promote trust between residents and officials, and enhance the adoption and acceptance of smart city technologies. However, there are still challenges to overcome, and continued research is necessary to fully realize the potential benefits of XAI.
Article
Full-text available
Different colleges and universities have different approaches to dealing with low-performance learners. However, in most cases, analgesics do not deal with root problems. This research suggests a model of three layers of variables sequentially adaptable to a deep-root issue. The suggested model can identify early pupils who could be at risk because of inaccurate or lack of match sequences and suggest rehabilitation. The approach proposed was implemented at three levels. First, we examined the personality type for 180 learners from different majors: Security and Forensics, Networking, and Application Development, using the MBTI test. Second, we build a knowledge matrix for courses by dividing each learning outcome into its knowledge segments. Then, we build the skills matrix for courses by decomposing each learning outcome into its skills segments. We then use machine learning (SVM, DT and association rules) algorithms to mine student performance on a smaller scale of knowledge and skills, taking into account their personality types instead of measuring an entire course's holistic performance. Finally, we developed a system of recommendations to detect performance deviations in knowledge and skills and provide adaptive learning materials that fit the examined students' personality. The proposed approach demonstrates its validity and effectiveness. However, it needs regular updates on learners' performance, which could be automated and linked to evaluation tools. The framework also has a minor impact on learners' privacy since it exposes individual personalities to their advisors.
Article
Full-text available
As smart metering technology evolves, power suppliers can make low-cost, low-risk estimation of customer-side power consumption by analyzing energy demand data collected in real-time. With advances network infrastructure, smart sensors, and various monitoring technologies, a standardized energy metering infrastructure, called advanced metering infrastructure (AMI), has been introduced and deployed to urban households to allow them to develop efficient power generation plans. Compared to traditional stochastic approaches for time-series data analysis, deep-learning methods have shown superior accuracy on many prediction applications. Because smart meters and infrastructure monitors produce a series of measurements over time, a large amount of data is accumulated, creating a large data stream, which takes much time from data generation to deployment of deep-learning model training. In this article, we propose an accelerated computing system that considers time-variant properties for accurate prediction of energy demand by processing the AMI stream data. The proposed system is a real-time training/inference system that deploys AMI data over a distributed edge cloud. It comprises two core components: an adaptive incremental learning solver and deep-learning acceleration with FPGA-GPU resource scheduling. An adaptive incremental learning scheme adjusts the batch/epoch in training iteration to reduce the time delay of the latest trained model, while trying to prevent biased-training due to the sub-optimal optimizer of incremental learning. In addition, a resource scheduling scheme manages various accelerator resources for accelerated deep-learning processing while minimizing the computational cost. The experimental results demonstrated that our method achieved good performance for adaptive batch size and epoch for incremental learning while guaranteeing a low inference error, a high model score, and queue stability with cost efficient processing.
Article
Full-text available
Explainability in Artificial Intelligence (AI) has been revived as a topic of active research by the need of conveying safety and trust to users in the “how” and “why” of automated decision‐making in different applications such as autonomous driving, medical diagnosis, or banking and finance. While explainability in AI has recently received significant attention, the origins of this line of work go back several decades to when AI systems were mainly developed as (knowledge‐based) expert systems. Since then, the definition, understanding, and implementation of explainability have been picked up in several lines of research work, namely, expert systems, machine learning, recommender systems, and in approaches to neural‐symbolic learning and reasoning, mostly happening during different periods of AI history. In this article, we present a historical perspective of Explainable Artificial Intelligence. We discuss how explainability was mainly conceived in the past, how it is understood in the present and, how it might be understood in the future. We conclude the article by proposing criteria for explanations that we believe will play a crucial role in the development of human‐understandable explainable systems. This article is categorized under: • Fundamental Concepts of Data and Knowledge > Explainable AI • Technologies > Artificial Intelligence Abstract Overview and discussion of different notions of explainability in approaches of expert systems, machine learning, recommender systems, and neuro‐symbolic learning and reasoning.
Conference Paper
Full-text available
Many recommendation systems suffer from overrun false rating injections that affect and shift generated recommendations in collaborative systems. Previous research focuses more on how to distinguish between profiles ignoring individual characteristics. This paper studies different behaviors and attributes of attackers and isolate them from genuine users. This paper suggested a new algorithm to filter two types of attacks, push and nuke attacks, where it evaluates users’ ratings according to the suggested algorithm and classify the rating as melt, freeze, or normal according to the calculate the variance from the center. The proposed algorithm gave high potentials and was able to filter such malicious attacks efficiently.
Article
Pandemics and natural disasters over the years have changed the behavior of people, which has had a tremendous impact on all life aspects. With the technologies available in each era, governments, organizations, and companies have used these technologies to track, control, and influence the behavior of individuals for a benefit. Nowadays, the use of the Internet of Things (IoT), cloud computing, and Artificial Intelligence (AI) have made it easier to track and change the behavior of users through changing the behavior of IoT devices. This article introduces and discusses the concept of the Internet of Behavior (IoB) and its integration with Explainable AI (XAI) techniques to provide trusted and evident experience in the process of changing IoT behavior to ultimately influence user behavior. Therefore, We propose a system based on IoB and XAI that aims to influence user behavior and showcase its effectiveness in an electrical power consumption use case. Through our experiments, we are able to demonstrate a reduction in both power consumption and cost. The scenario results showed a decrease of 522.2 kW of active power based on the original historical average consumption over a 200-hours period. It also showed a total power cost saving of C95.04 for the same period.
Article
Machine learning algorithms have been used to predict student performance in academic institutions over the last decade. The developed prediction models classified students as either those who were likely to receive a distinctive or those who were likely to be at-risk or withdraw from the class. Although, the early prediction aid in the targeting of educational interventions and achieved a more efficient allocation of educational resources. All proposed solutions fell within the scope of predictions that result in active or proactive actions to support universities and learners. On the other hand, they fail to comprehend the various forms of education systems and whether it appropriate for the twenty-first century and future generations. The paper classifies education into five types based on the design mode, the scope of production, and the interaction between learners and educational systems (Push, Pull, coupling, Integrated, and Sustainable). The paper proposes a sustainable education paradigm that maximizes the knowledge and skill matrix accumulated for the desired program. The proposed theory implementation phases are modelled and demonstrated using 21st-century technologies, such as personalized and coaching education based on the learner’s learning style and remediation actions for strong learners with innovative competencies. The study emphasized various aspects of sustainable education systems that are required for smart city transition. The limitations and proposed solutions for dealing with anticipated issues were demonstrated, and the benefits of sustainable education are based on the proposed maximization theory rather than the current block-based learning of outcomes.