See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/381322513

Digital Intelligence for University Students Using Artificial Intelligence

Techniques

Preprint in International Journal of Computing and Digital Systems · January 2025

DOI: 10.12785/ijcds/1571029446

CITATIONS READS
0 164

5 authors, including:

Ban Hassan Majeed Rasha H. Ali

University of Baghdad; College of Education for Pure Science/ Ibn Al-Haitham University of Baghdad
70 PUBLICATIONS 734 CITATIONS 19 PUBLICATIONS 64 CITATIONS

SEE PROFILE SEE PROFILE

Shahlaa Mashhadani
University of Baghdad
16 PUBLICATIONS 15 CITATIONS

SEE PROFILE

All content following this page was uploaded by Shahlaa Mashhadani on 02 January 2025.

The user has requested enhancement of the downloaded file.

 International Journal of Computing and Digital Systems
2025, VOL. 17, NO.1, 1-10
http://dx.doi.org/10.12785/ijcds/1571029446

Digital Intelligence for University Students Using Artificial

Intelligence Techniques

Ban Hassan Majeed1 , Wisal Hashim Abdulsalam2 , Zainab Hazim Ibrahim3 , Rasha H. Ali4 and Shahlaa
Mashhadani5
1,2,3,5
Computer Department, College of Education for Pure Science/Ibn-Al-Haitham, University of Baghdad, Baghdad, Iraq
4
Computer Science Department, College of Education for Women, University of Baghdad, Baghdad, Iraq

Received 11 May 2024, Revised 11 November 2024, Accepted 28 November 2024

Abstract: The research problem arose from the researchers’ sense of the importance of Digital Intelligence (DI), as it is a basic
requirement to help students engage in the digital world and be disciplined in using technology and digital techniques, as students’
ideas are sufficiently susceptible to influence at this stage in light of modern technology. The research aims to determine the level of
DI among university students using Artificial Intelligence (AI) techniques. To verify this, the researchers built a measure of DI. The
measure in its final form consisted of (24) items distributed among (8) main skills, and the validity and reliability of the tool were
confirmed. It was applied to a sample of 139 male and female students who were chosen in a random stratified manner from students at
the University of Baghdad, College of Education for Pure Sciences/Ibn Al-Haitham, Department of Computer. The proposed AI model
utilized three artificial intelligence techniques: Decision Tree (DT), Random Forest (RF), and Gradient Boosting Machine (GBM). The
classification accuracy using DT was 92.85 and using GMB was 95.23. The RF technique was applied to find the essential features,
and the Pearson correlation was used to find the correlation between the features. The findings indicated that students indeed possess
digital intelligence, underscoring the potential for tailored interventions to enhance their digital skills and competencies. This research
not only sheds light on the current DI landscape among university students but also paves the way for targeted educational initiatives
to foster digital literacy and proficiency in the academic setting.

Keywords: Digital Intelligence, Artificial Intelligence, University Students, DT, RF, GBM, Pearson correlation.

1. Introduction tutions, ministries, and companies in almost all industries.

Digital intelligence (DI) enhances innovation and cre- AI can help accomplish many professional, functional, and
ativity capabilities through technology and digital tools, as commercial activities with greater accuracy and in a small
individuals can transform their ideas into reality and develop fraction of the time compared to what it takes humans to do.
new solutions to complex problems. This requires thinking It also provides an alternative to human judgment because
innovatively and creatively using digital tools [1]. AI can analyze and predict a lot of data based on available
However, we must face some challenges associated with data that humans cannot easily discover, as demonstrated in
DI. For example, overreliance on technology can lead to references [2], [3], [4].
a decline in basic skills such as concentration and critical The importance of DI, is that it is a social, cognitive,
thinking. We may also have to face challenges related to and emotional ability that enables the individual to face
digital privacy and security. Therefore, we must be careful the challenges and situations that he/she is exposed to
and learn to use technology safely and responsibly. in the digital world in a proper way and which gives
Artificial intelligence (AI) is rapidly emerging as the most the individual appropriate and responsible behavior during
important transformative technology of our time and is his/her interaction in the digital environment [5].
considered one of the biggest challenges facing organiza- The contribution and importance lie at two levels, theoreti-
tions seeking to transform into the digital age, as recent cal and applied. The theoretical importance comes from the
developments, especially in machine learning (ML), have absence of any study that measures DI using AI techniques
led to the computer’s ability to improve its performance for university students in Arab countries, especially Iraq,
without human intervention, with the rapid spread of new to the best of researchers’ knowledge. This research is also
applications that have changed the work laws of many insti- important because it sheds light on DI, its skills, and the

E-mail address: ban.h.m@ihcoedu.uobaghdad.edu.iq, wisal.h@ihcoedu.uobaghdad.edu.iq, zainab.h.i@ihcoedu.uobaghdad.edu.iq,

rashaha2003@coeduw.uobaghdad.edu.iq, shahlaa.t@ihcoedu.uobaghdad.edu.iq
2 Ban Hassan Majeed, et al.

importance of the role of university institutions in providing 2. Related Works

their students with these skills, and because technology, AI, A. Digital Intelligence (DI)
and the products and services they provide are important in It is the set of technical, cognitive, social, and emotional
the state’s development plans in all fields. Education has a capabilities that enable people to use digital tools efficiently,
particularly large share, which demonstrates the importance and they have the right to know their digital identity and
of technology. Furthermore, with AI and its entry into our their rights and duties in using it, dealing with digital
lives and with this rapid growth in technology, many enor- problems, and adapting to the requirements of digital life
mous ethical, educational, and technical challenges have in a way that achieves safety and security for them [6]. It
emerged, and the use of AI, which is expected to result in is also the ability to understand digital concepts through
unprecedented social phenomena, calls for the importance the Internet and use them appropriately, and to solve tech-
of thinking about the uses and applications of intelligence nological and information problems, and communication
and that the needs of today’s education are best met through problems of communication via the Internet [7]. It is defined
the concept of developing DI that includes various areas and procedurally as the amount of score that students obtain
many skills. Also, the importance of this research stems when dealing with the items on the scale. It represents
from the fact that the researchers developed a scale to the ability of individuals to deal with knowledge and use
measure its availability among students of a college with a digital technologies in a way that enables them to interact
scientific specialization within a government university that successfully and usefully with their digital environment.
is considered the mother university for Iraqi universities in Their capabilities are divided into three levels as follows
Iraq. [8]:
Enriching educational content with DI and its skills, which
are considered necessary skills in light of the technological 1- Digital citizenship refers to the ability to use tech-
developments and progress we are witnessing in our current nology and digital media in a safe, responsible, and
era, and highlighting DI skills and the necessity of acquiring moral way.
them for our students, because these skills will help students 2- Digital creativity refers to the ability to become part
in the responsible, safe, and ethical use of digital space. of the digital ecosystem.
It helps curriculum developers recognize the importance 3- Digital competitiveness refers to the ability to solve
of including DI skills in content. Teaching and activities global challenges, innovate, and create new opportu-
appropriate to the characteristics of the stage. The research nities in the digital economy by stimulating increased
results present the level of possession of DI by students, business, jobs, growth, and impact.
which contributes to developing plans and programs to
develop it and enhance their skills. There is also a severe B. Skills of DI
lack of educational research and studies that have dealt with DI emerges from human values, such as respect, ac-
DI, as well as the computerization of a measure of DI. cepting differences, activating positive communication, and
This research is considered one of the first in Iraq, as DI, other things that enable the student to acquire DI skills
as we are witnessing, is considered one of the important and to be a leader in technology, not a coordinator of it.
goals in the educational and pedagogical process in the Teaching DI skills and imparting them to our students has
current era. Due to the technical revolution represented become an urgent necessity in light of the developments
by AI products and services and the involvement of all we are experiencing in technology [9], the fourth industrial
members of society using it. The importance of DI has revolution, AI, and the Internet of Things (IoT) knowing
become a pivotal role in shaping the personality of the that these technologies have been in the last few years.
university student and directing his/her mental abilities and The future will be an essential part of everyone’s life, and
practical practices toward achieving a better reality through therefore students need these skills to have a better future
the optimal use of his/her mental and intellectual abilities. and opportunities:
In addition, providing a computerized scale to measure DI
among university students that can be used. This research 1- Screen Time Management (STM): It is the time spent
is also important because it sheds light on DI skills, which using any device, whether it is a TV, A computer,
prepare the individual to deal with technology in the digital a smartphone, video games, or a tablet for enter-
world with sufficient awareness by improving and benefiting tainment [10] It is the ability to manage time when
from the positives and preventing and avoiding risks. The using the screen. This skill aims to raise awareness
practical importance comes from providing a computerized among student users of the harms caused by pro-
measure of DI for university students that other researchers longed screen time, such as electronic addiction and
can benefit from. disruptive behavior, as well as the neglect of other
Our limitations are: responsibilities. This skill also enables the student to
master the balance between screen time and his/her
1) The students are from all the 4th grades (1-4) in the other tasks outside the Internet. It also determines
Computer Department in the College of Education. his/her priorities, the need for parents to participate
2) 1st Semester 2023-2024. in Internet activities, and their awareness of them
3) DI Test. [11].
 International Journal of Computing and Digital Systems 3

2- Digital footprint management (DFM): The digital tention to them, and respond to them. Friends and
fingerprint is the data that is formed during the use of acquaintances on social networking sites, respect-
the Internet, including the websites the student visits ing the point of view of others who disagree with
and logs in to, the emails, photos, and videos that their point of view and their opinion, confronting
he/she sends [12] indicated that the digital footprint speeches, reducing hatred and violence, and control-
includes activity in the digital environment and that ling emotions when exposed to emotional situations
what is published on the Internet lasts for years; electronically [7].
even if it is deleted, it can be retrieved. Therefore, 6- Digital Citizen Identity (DZI): [16], [17] Defined
students must be creative and positive in using online digital citizenship as a code of conduct concerning
content, understanding the benefits of cooperative the optimal use of technology. Digital citizenship is
and group learning using Internet tools, sharing important among learners, training them practically
knowledge and resources, and focusing on protecting in dealing with modern digital technologies and pro-
their personal information and building a positive viding them with the appropriate means and methods
digital reputation. This skill aims at allowing the to take the proper solution if they are exposed to
student to understand the nature of communication electronic blackmail.
via the Internet, to raise his/her level of responsibility 7- Electronic (Cyber) Security Management (ECM):
when he/she is in contact with the digital world, Cybersecurity is defined as targeting websites
and to be sufficiently aware of the enrichment of through other electronic means and is a group of
his/her digital imprint, whether positive or negative, practices that aims to protect systems, networks,
on his/her digital reputation. The student must realize programs, and individuals from digital attacks of
that this digital imprint is permanent. They cannot be any kind. These various practices have proactive
deleted or changed in the digital space, indicating precautionary measures before the failure occurs and
that the digital footprint includes activity in the remedial measures after the failure occurs. So, it is
digital environment and that what is published on a digital skill that enables an individual to create
the Internet lasts for years; even if it is deleted, word traffic and be strong, which will contribute
it can be retrieved. Therefore, students must be to avoiding electronic attacks. This skill aims to
creative and positive in using online content and empower the individuals to protect themselves from
understanding the benefits of cooperative and group the dangers of the Internet from phishing, intrusion,
learning using Internet tools, sharing knowledge and and theft of his/her data and personal information
resources, as well as focusing on protecting their [18].
personal information and building a positive digital 8- Critical Thinking: Defined by [19] as the type of
reputation [13]. thinking that depends on careful examination of all
3- Management of privacy (MP): It is an important dig- premises and evidence, then is based on a slow step-
ital skill that enables an individual to protect his/her by-step progression and is characterized by objec-
information, personal data, and contact information tivity to the greatest extent possible with the aim
from theft, plagiarism, etc. This skill aims to raise of arriving at sound results characterized by validity,
awareness of the danger of sharing their data and stability, and honesty. It is one of the most important
their personal information and realize the importance types of thinking that must be paid attention to
of maintaining privacy on social media sites. The by those in charge of education in general, as the
student also learns to handle his/her information and learner can distinguish between what is good from
disclose it, especially since social networks collect the information and ideas he/she receives and not
users’ personal information, and from this informa- accept any customs and traditions, whether inherited
tion and data, they can direct products, services, etc. or imported, except after reconsidering them and
to these users. This is the type of targeted advertising making judgments. This skill makes the student pos-
from which social media generates huge profits [14]. itively critical in distinguishing between information
4- Managing Cyberbullying (MC): It requires a nec- and content. The real and false, between good and
essary skill for students to detect cases of bullying bad friendship relationships, and safe and unsafe
and bullying on the Internet and deal with them con- activities, and how to beware of strangers in the
sciously and cautiously to avoid harm. The parties to digital world [20].
cyberbullying are the bully, the bullied (the victim),
the supporter of the bully, and the silent, passive Researchers believe that these skills are a comprehensive
person who does not accept bullying but is silent set of technical, cognitive, social, and emotional skills that
and ignores it [15]. enable individuals and societies to face challenges and
5- Digital empathy (DE): Develop relationships in the harness digital life, as a tool was built to measure DI based
digital community, pay attention to the feelings of on these main skills.
others, and build good relationships in the online
world. It is defined as the individual’s ability to
consider the needs and feelings of others, pay at-
4 Ban Hassan Majeed, et al.

C. The importance of DI lies in [21] 6- Human interaction ability: The ability to interact and
1- Enables students to face and adapt to the demands communicate with humans naturally and effectively.
of digital life.
2- Manage student digital identity online and become C. The Importance of AI
aware of it. 1- The efficacy and effectiveness of AI; it is used with
3- The ability to use and control digital devices to robotics in the mining and exploration process.
achieve a balance between life online and outside. 2- Digital assistance programs: Help us in daily work.
4- The ability to manage risks and detect electronic For example, there are personal assistance programs
threats. that listen to us perform the task with one click and
5- Communicate, empathize with others via the Inter- help us travel around the world.
net, and build good relationships with them. 3- Helping to make rational and correct decisions: For
6- Create and evaluate online content. AI, logic has priority over the rest of the components,
7- Understanding personal and legal rights and freedom as it makes correct decisions if they are worked
of expression online. on rationally, but we see the opposite with humans,
where at some point they use emotion. Robots or ar-
3. Artificial Intelligence (AI) tificial thinkers have no side. Emotional intelligence
Is a computer science field related to computer sys- is what makes people more effective, as it can make
tems that possess characteristics associated with intelli- the right decisions in a short period.
gence, decision-making, and similarity to human behavior 4- Medical applications: Among the most important
in various fields [22]. It is the result of the contemporary uses of AI in the medical field are those programs
technological revolution. AI initially aimed to simulate each that are directed at assessing patients and their health
of the various intelligence capabilities through machines by risks with the help of machines [26].
understanding the complex mental processes that the human 5- Eliminating repetition in performing tasks: It is
mind performs [23]. useful in eliminating repetition of tasks resulting
from the multiplicity and frequency of tasks and the
A. Characteristics of AI person being under a set of pressures, as it includes
Among its characteristics is the ability for [24]: preparing specific programs for tasks, to avoid repeti-
tion or randomness in executing the orders entrusted
1- Thinking and perception. to him, given its distinctive features. These programs
2- Discovering and applying knowledge. are complex and organize the work that must be
3- Learning and understanding from previous experi- carried out in the form of a schedule that eliminates
ences and expertise and using them in new situations. the possibility of repetition.
4- Using trial and error to explore disputed matters. 6- Time management: A large number of companies
5- Rapid response to new situations and circumstances. and institutions are under pressure resulting from
6- Development, creativity, understanding, and percep- mismanagement of time and repeatedly exploiting it
tion of visual matters. to perform the same tasks. In the same context, AI
7- Providing information to support administrative de- works to save a lot of time for these companies by
cisions. inventorying all data and collecting it in less than a
8- Finally, it is used to solve problems presented in the period. Man saves time from wasting.
absence of complete information. 7- The possibility of representing knowledge: AI pro-
grams use a special structure to describe knowledge,
B. Features of AI [25] and this structure includes facts and relationships
These features highlight the benefits of AI in various linked to each other by rules, most notably the
fields, such as analyzing large amounts of data, developing knowledge base, which provides access to a large
innovative interactive applications, and solving complex volume of information about The problem to be
problems. solved with great ability to represent it.
8- Delving deeper into specializations: AI can help in
1- Processing speed: the ability to process data and in-depth research of users’ desires and preferences,
perform calculations quickly and effectively. helping to identify individuals’ tastes and requests,
2- Accuracy: the ability to perform tasks with high working to meet and implement them, and creating a
precision and obtain accurate results. complete vision of the required and desired product.
3- Continuous learning: The ability to learn from pre- This is one of the most important methods used in
vious data and experiences to enhance performance advertising and marketing products.
over time. 9- Software and humans at the same level: When AI
4- Dealing with large information: the ability to deal is introduced on a large scale, software and humans
with, analyze, and extract information from large sets will become like one team at work, as applications
of data. will be dealt with through software and not humans
5- Diagnosis and problem solving. [27].
 International Journal of Computing and Digital Systems 5

To the researchers’ knowledge, this is the first study of

its kind, as it included a sample of students from a college
with a scientific specialization in the field of education and
upbringing. This is considered positive for the researchers,
as they are academic professors who used their students to
measure a numerical variable.
4. Materials and Methods
This study used baseline data from the students’ survey
of stages (1-4) in the Computer Department / College of
Education for Pure Science, Ibn-Al-Haitham / University
of Baghdad, IRAQ. One hundred and thirty- nine students
consented to participate.
Classification and Feature Selection
An overview of the feature selection and classification
methods used is provided below.
Feature selection is a fundamental part of pattern recogni- Figure 1. The proposed model architecture
tion and ML, as it removes irrelevant, noisy, and extraneous
data from raw features, which leads to an improvement in
the effectiveness of prediction models. In this study, feature 5. The Proposed Model
importance scores were derived with the Random Forest In general, the idea of the suggested model is to
(RF) algorithm, which is employed in ML [28]. Builds create a model using the Decision Tree (DT) algorithm
several decision trees; each tree uses the Gini index (or and Gradient Boosting Machine (GBM), which is one
another) to find the best way to split the data, favoring of the ML algorithms used in classifying the features of
splits that maximize class separation. The final decision is the dataset. This suggested model contains four stages:
determined by the majority vote of all individual trees. RF data collection, pre-processing, feature construction, and
has many advantages, such as being able to handle complex classification. As shown in Fig. 1, feature selection has
data well, robust to outliers, and easier to interpret [29]. been used using the Random Forest algorithm, which is
A Decision Tree (DT), a widely used method in machine one of the machine learning algorithms; this algorithm is
learning, is considered highly valuable due to its reason- used to select important features that affect classification,
able accuracy and relatively low computational cost. This and the correlation between features has been calculated
supervised learning technique classifies labeled training data using Pearson’s correlation.
into a tree or set of rules. Recursively divides a data domain
(node) into two subdomains based on the highest informa-
tion gain to optimize the split for maximum information
gain [30].
The Gradient Boosting Machine (GBM) is a powerful
machine learning algorithm that is gaining considerable
traction in a wide range of data-driven applications, such
as computer vision [31].
A. Correlation Coefficient
Pearson is a statistical indicator of how much the
values of two variables vary from each other and shows
how strongly two variables are linearly connected. The
correlation coefficient, which measures the linear correlation
between two discrete-time signals, is defined in (1) [32]: Figure 2. The histogram of male and female counting
Cov x,y
P x,y = (1)
σ x σy A. Data Collection
where x and y represent the standard deviations of x[n] Data were collected using a survey, which was com-
and y[n], respectively, and Covx, y denotes the covariance pleted by the undergraduate students of stages (1-4) in
of two signals. The correlation coefficient, x, y, ranges in which they are a research sample. The questionnaire con-
value from −1 to +1. There is no linear relationship between tained 12 groups with 24 questions, which were taken
the two signals if the value is equal to zero. A significant up by the specialists. The questionnaire statements were
positive correlation between the two signals corresponds to written in Arabic and distributed to students through Google
a value near +1 for the x, y values, and a strong negative Classroom. The students had answered only one hundred
correlation to a value close to −1.
6 Ban Hassan Majeed, et al.

TABLE I. THE DIGITAL VALUE OF THE ATTRIBUTES

Attribute Scores Based on Likert Scale

I strongly agree 5
I agree 4
Neutral 3
I disagree 2
I strongly disagree 1

thirty-nine. Most of the responses are from males. Fig. 2

shows the male and female histograms in the dataset.
B. The Pre-processing Stage
This pre-processing step is vital in the prediction pro-
cess, as it repairs the data to ensure that it can be employed
in feature selection and hence the classification stage. Two
processes were carried out at this stage. First, the discretiza-
tion of attributes was conducted. In the questionnaire, the
values of each attribute were divided according to “The
Likert scale.” The value of scaling digitally is presented
in TABLE I. The class label classifies binary classification
as either existing or not. The class label was computed by
the threshold using Eq. (2) for each instance. In this work,
the threshold value was (95). Secondly, processing the null
values (if there are any) in the dataset by taking the mean
using Eq. (3). The dataset contains 139 records. The dataset Figure 3. The histogram of the class label for databaset
contained ( 87 for one-class labels and 52 for zero-class
labels). Fig. 3 shows the histogram of the class label.
X max x + min x Recall is the rate of correctly predicted positive observations
S = (x); T = (2) for all observations in an actual class, as shown in Eq. (5)
2
below:
N TP
µ=
1 X Recall = (5)
N i=1
xi (3) T P + FN
Precision is the ratio of correctly predicted positive observa-
tions to the total predicted positive observations, as shown
C. Features Construction Stage in Eq. (6) below:
At this stage, the attributes are directly selected using the TP
feature selection method to build the artificial intelligence Precision = (6)
T P + FP
model. There were 24 attributes in the dataset. Using the
questionnaires indicated in the previous section, the students The F1 score is a measure of the accuracy of a test.
filled out these characteristics. The score of each feature Considers both the precision and the recall of the test to
ranges from 1 to 5. TABLE ?? presents the feature number compute the score, as appears in Eq. (7) below:
in a model of artificial intelligence, attribute, and description
of the feature, while TABLE II shows the group of features. 2 ∗ Recall
F1 − score = (7)
D. Evaluation of the Classification Stage and Performance Precision + Recall
As mentioned in the previous subsection, the feature
E. The Classification Stage
vector was constructed from the responses to the question-
naire. The number of features is twenty-four. DT technique In this stage, the model is created using a Decision
and GBM have been used for classifying the feature vector. Tree (DT) and Gradient Boosting Machine (GBM), and the
The dataset contained 139 instances. The proposed work is dataset is split for training the model (70 %) and for testing
evaluated using the following equations [33] (30%). The DT and GBM are used to classify the dataset.
The model is trained using training data. By this step, the
TP + TN model learns the patterns of features and relationships in
Accuracy = (4)
T P + FN + T N + FP the data to class labels.
where T P = true positive, FN = f alse negative, FP =
f alse positive, and T N = true negative.
International Journal of Computing and Digital Systems 7
8 Ban Hassan Majeed, et al.

TABLE II. THE GROUPS OF THE FEATURES TABLE III. THE SCORE OF EACH ATTRIBUTE USING RF.

Feature No. Feature description Feature Feature Feature Feature

F0-F2 Controlling Screen Time No. score No. score
F3-F7 Managing Social Network F0 0.016098 F12 0.02504
F8-F9 Protecting My Rights F1 0.011631 F13 0.05382
F10-F11 Dealing with Harmful Content F2 0.031149 F14 0.02022
F12-F14 Building Positive Relationships F3 0.021091 F15 0.06315
F15-F16 Using Technology Responsibly F4 0.029370 F16 0.01930
F17 Digital Leadership F5 0.038870 F17 0.05064
F18-F19 Data Protection F6 0.022486 F18 0.03106
F20 Cybersecurity F7 0.023642 F19 0.18191
F21 Digital Literacy F8 0.013950 F20 0.02274
F22 Distinguishing Reliable Sources F9 0.00926 F21 0.04025
F23 Evaluating Digital Content F10 0.07474 F22 0.05202
F11 0.11674 F23 0.03074

Figure 4. The Sorting of features depending on importance using Figure 5. The Pearson correlation between features
RF

features and to find out whether the attribute depends on

F. The Importance of Features one or multiple features. The correlation was calculated
The Random Forest (RF) was used to locate the impor- using Eq. (1). The positive correlation between two or more
tance of features by computing the Gini impurity for each than one feature, the positive correlation value greater than
feature. Therefore, each feature had a score and weight, zero, which represents a slight cell in Figure 5. A positive
while the gain type shows the average Gini impurity across correlation means that any increase in the value of one
all splits where the feature was used, and the weight shows feature causes an increase in the value of the correlated
the number of times the feature is used to split the data. feature. As shown in Fig 5 there is more than one feature
Fig. 5 shows the histogram of important features using RF, that is correlated: feature (F8) The description is ”Con-
while Table IIII shows the importance of each feature. trol the personal information and data shared online with
confidentiality.” more correlated with feature (F20), whose
As shown in Fig. 4 and TABLE III, the feature ”Use description is ”Use appropriate security tools to protect
appropriate security tools to protect data against any cyber- data against any cyberattack, such as camera closing.”. The
attack, such as camera closing,” has the highest degree in feature (F23) that the description is (”I can critique digital
the important features using RF technique, and the feature content correctly and positively”) is more correlated with
”Avoid risks such as content and communication risks” feature (19), which is the description (”Use appropriate
has the highest degree. The features ”Detect and deal with security tools to protect data against any cyberattack, such
cyberbullying cases with caution” and ”Control screen time as camera closing”). Also, the feature (”Control screen
when participating in online games” have the lowest degree, time when using any electronic screen”) correlated with
respectively. This means that each feature has a degree that the feature (”Use appropriate security tools to protect data
affects the prediction of DI. against any cyber attack, such as camera closing”).
G. Correlation of Features 6. Results and Discussion
The correlation between features was computed using This work was carried out using the Python (V.3.9) pro-
Pearson’s correlation to understand the relationship between gramming language and the Jet Brains Pycharm (V.2018.2)
 International Journal of Computing and Digital Systems 9

TABLE IV. THE CLASSIFICATION REPORT OF DT.

academic teachers in educational institutions dedicated to
developing their digital intelligence skills, which would

Class label
Technique

Precision

Support
Recall
increase and enhance their student’s ability to think with

FI
precise digital skills consistent with the development taking
place in the technological field. In the future, a study will
0.79 1.00 0.88 11 0 be done to determine the relationship between digital intel-
DT 1.00 0.90 0.95 31 1 ligence and strategic intelligence among university students.
0.93 42 Accuracy In addition, we suggest performing a comparative study
0.85 1.00 0.92 11 0 between public and private universities by adopting a large
GBM 1.00 0.94 0.97 31 1 sample and using deep learning methods to generalize the
0.95 42 Accuracy results.
References
[1] S. S. Hammadi, B. H. Majeed, and H. A. K., “Impact of deep learn-
framework. As mentioned, this work contained four main ing strategy in mathematics achievement and practical intelligence
stages. Firstly, data collection by answering the survey. among high school students,” Int. J. Emerg. Technol. Learn., vol. 18,
Secondly, data preprocessing by processing the null value. pp. 42–52, 2023.
Thirdly, construct a feature vector by selecting all features
for classification in the next stage. Finally, the classification [2] R. H. Ali and W. H. Abdulsalam, “The prediction of covid 19
stage was carried out using the DT and GBM. The DT and disease using feature selection techniques,” Journal of Physics:
Conference Series, vol. 18, p. 022083, 2021.
GBM algorithms were applied to the dataset to predict the
digital intelligence of university students. The classification [3] W. H. Abdulsalam, R. S. Alhamdani, and M. N. Abdullah, “Emo-
accuracy was 92.85 % and 95.23,% respectively. TABLE tion recognition system based on hybrid techniques,” International
IV shows the classification report for the DT classifier. Journal of Machine Learning and Computing, vol. 9, pp. 490–495,
2019.
As shown in Table IV, the classification report for RF
and GBM was measured to clarify the results, especially [4] R. H. Ali and W. H. Abdulsalam, “Attention-deficit hyperactivity
in the imbalance of the dataset. Generally, the value of disorder prediction by artificial intelligence techniques,” Iraqi Jour-
nal of Science, vol. 65, pp. 5281–5294, 2024.
F1 score was 0.95% and 0.97, % respectively, for the class
label (one) and 0.88% and 0.92, % respectively, for the class [5] M. R. Yousif, L. T. Ameen, B. M. Jassim, and B. H. Majeed,
label (zero). The values of the accuracy and F1 score were “The impact of two proposed strategies based on active learning on
biased for the class label (one) because the number of class students’ achievement at the computer and their social intelligence,”
labels (one) is higher than the class label (zero). As shown International Journal of Engineering Pedagogy, vol. 14, 2024.
in the results, the proposed model using DT and GBM was
[6] M. S. K. Wahib, Z. A. A. Alamiry, B. H. Majeed, and H. T.
successful in predicting the digital intelligence of university H. S. Alrikabi, “Digital citizenship for faculty of iraqi universities,”
students by training the factors of digital intelligence using Periodicals of Engineering and Natural Sciences, vol. 11, pp. 262–
a machine learning algorithm (DT) and (GBM). Also, by 274, 2023.
using the machine learning algorithm (RF) for computing
the importance of features that affect the decision, measure [7] D.-M. Cismaru, P. Gazzola, R. S. Ciochina, and C. Leovaridis, “The
digital intelligence. rise of digital intelligence: challenges for public relations education
and practices,” Kybernetes, vol. 47, pp. 1924–1940, 2018.
7. The Conclusions
[8] D. K. A.-R. Al-Malah, B. H. Majeed, and H. T. H. S. Alrikabi,
In the current digital age, DI is essential for individuals “Enhancement the educational technology by using 5g networks,”
and organizations. Technology is used to achieve goals and International Journal of Emerging Technologies in Learning (On-
solve problems. Recently, AI techniques in general, have line), vol. 18, p. 137, 2023.
been used for predicting behavioral and social diseases, as
well as medical diseases. So, the current study presented [9] B. Hamilton, “Integrating technology in the classroom: Tools to
meet the needs of every student,” International Society for Technol-
the prediction of the digital measure for university stu- ogy in Education, 2022.
dents using AI techniques. One hundred thirty-nine students
completed an online survey that contains the statements of [10] L. S. George and C. L. Park, “Meaning in life as comprehension,
digital measures, numbered forty-four features to investigate purpose, and mattering: Toward integration and new research ques-
the degree of benefit or damage. The proposed model tions,” Review of general psychology, vol. 20, pp. 205–220, 2016.
used three AI techniques: decision tree, gradient boosting
machine, and random forest. The classification accuracy [11] M. Shanmugasundaram and A. Tamilarasu, “The impact of digital
technology, social media, and artificial intelligence on cognitive
using DT and GBM was 92.85 and 95.23, respectively. functions: a review,” Frontiers in Cognition, vol. 2, p. 1203077,
The RF technique was applied to find the essential features 2023.
and the Pearson correlation to find the correlation between
features. [12] A. K. Jain, S. R. Sahoo, and J. Kaubiyal, “Online social networks
The researchers recommend holding training courses for
 10 Ban Hassan Majeed, et al.

security and privacy: comprehensive review and analysis,” Complex [23] J. Sublime, “The ai race: Why current neural network-based archi-
& Intelligent Systems, vol. 7, pp. 2157–2177, 2021. tectures are a poor basis for artificial general intelligence,” Journal
of Artificial Intelligence Research, vol. 79, pp. 41–67, 2024.
[13] S. Venkadasubbiah, D. Yuvaraj, S. Ali, and M. U. A. Ayoobkhan,
“Data footprinting in big data,” Big Data Analytics and Computa- [24] G. F. Luger and W. A. Stubblefield, “Ai algorithms, data structures,
tional Intelligence for Cybersecurity, vol. 18, pp. 203–218, 2022. and idioms in prolog, lisp, and java: Pearson addison-wesley,” 2009.

[14] J. Puhakka, “Challenges of targeted ads in social media,” 2021. [25] B. Pliuskuvienė, U. Radvilaitė, P. Stefanovič, and S. Ramanauskaitė,
“Machine learning-based chatgpt usage detection in open-ended
[15] E. V. Ugochukwu, I. A. Okenwa-Fadele, M. N. Nwikpo, U. I. question answers,” 2023.
Oparaugo, and C. C. Okoro, “Students’ and teachers’ perception
of cyberbullying and its influence on the academic grit of senior [26] S. Sabeeh and I. S. Al-Furati, “Issues and research fields of medical
secondary school students in onitsha south lga, anambra state,” robotics: A review,” Iraqi Journal for Electrical & Electronic
Asian Journal of Education and Social Studies, pp. 28–38, 2022. Engineering, vol. 19, 2023.

[16] M. S. Ribble, G. D. Bailey, and T. W. Ross, “Digital citizenship: [27] W. H. Abdulsalam, S. Mashhadani, S. S. Hussein, and A. A. Hashim,
Addressing appropriate technology behavior,” Learning & Leading
“Artificial intelligence techniques to identify individuals through
with technology, vol. 32, p. 6, 2004.
palm image recognition,” International Journal of Mathematics and
Computer Science, vol. 20, pp. 165–171, 2025.
[17] Z. H. Ibrahim, “Analysis of computer textbook for the second
intermediate grade according to digital citizenship,” Opción: Revista [28] X. Sang, W. Xiao, H. Zheng, Y. Yang, and T. Liu, “Hmmpred:
de Ciencias Humanas y Sociales, vol. 28, 2019. accurate prediction of dna-binding proteins based on hmm profiles
and xgboost feature selection,” Computational and mathematical
[18] W. A. S. . Z. M. J. K. . S. S. Ahmed, “Novel standard polynomial as methods in medicine, pp. 1748–670X, 2020.
new mathematical basis for digital information encryption process,”
Advances in Decision Sciences, vol. 27.
[29] W. H. Abdulsalam, R. S. Alhamdani, and M. N. Abdullah, “Speech
emotion recognition using minimum extracted features,” 1st Annual
[19] B. H. Majeed, L. F. Jawad, and H. AlRikabi, “Tactical thinking and International Conference on Information and Sciences (AiCIS) Fal-
its relationship with solving mathematical problems among math- lujah, Iraq, pp. 58–61, 2018.
ematics department students,” International Journal of Emerging
Technologies in Learning (iJET), vol. 16, pp. 247–262, 2021.
[30] T. B. Alhijaj, S. M. Hameed, and A. A. Bara’a, “A decision tree-
aware genetic algorithm for botnet detection,” Iraqi Journal of
[20] B. H. Majeed, “Impact of a proposed strategy according to luria’s Science, pp. 2454–2462, 2021.
model in realistic thinking and achievement in mathematics,” Int. J.
Emerg. Technol. Learn., vol. 17, pp. 208–218, 2022.
[31] S. Touzani, J. Granderson, and S. Fernandes, “Gradient boosting
machine for modeling the energy consumption of commercial build-
[21] H. T. H. Salim, A. L. Rikabi, M. J. Al-Dujaili, B. H. Majeed,
ings,” Energy and Buildings, vol. 158, pp. 1533–1543, 2018.
and I. R. N. Alrubeei, “Information and communication technology
and its impact on improving the quality of engineering education
systems,” International Journal of Engineering Pedagogy, vol. 14, [32] H. Bakhshayesh, S. P. Fitzgibbon, A. S. Janani, T. S. Grummett,
2024. and K. J. Pope, “Detecting synchrony in eeg: A comparative study
of functional connectivity measures,” Computers in biology and
[22] W. H. Abdulsalam, R. S. Alhamdani, and M. N. Abdullah, “Facial medicine, vol. 105, pp. 1–15, 2019.
emotion recognition from videos using deep convolutional neural
networks,” International Journal of Machine Learning and Com- [33] J. Davis and M. Goadrich, “The relationship between precision-
puting, vol. 9, pp. 14–19, 2019. recall and roc curves,” pp. 233–240, 2006.

View publication stats