Intelligent Tutoring System for Computer Science Education and the Use

of Artiﬁcial Intelligence: A Literature Review

Rodrigo Elias Francisco

1,2 a

and Flávio de Oliveira Silva

1 b

Faculty of Computer, Federal University of Uberlândia (UFU),

Av. João Naves de Ávila, 2121, Block 1A, Room 1A243 - Campus Santa Mônica, Uberlândia - MG, Brazil

Federal Institute Goiano (IF Goiano) - Campus Morrinhos, Rodovia BR153, KM633 Zona Rural, Morrinhos - GO, Brazil

Keywords:

Intelligent Tutoring System, Computer Science Education, Artiﬁcial Intelligence, Literature Review.

Abstract:

Education in computer science brings speciﬁc challenges to the teaching-learning process. Students spend

a lot of time dealing with the complexity of problems and learning to use existing technologies. Intelligent

Tutoring System (ITS) is a technology that can contribute to this scenario, automating and adapting teaching

to the student’s proﬁle. This work presents a literature review on ITS’s for Computer Science Education,

focusing on Artiﬁcial Intelligence (AI) in this scenario. We analyze the development and use of ITS’s in

Computer Science Education and assess AI techniques, algorithms, and datasets. The results of this review

point to challenges in research on aspects such as the unavailability and difﬁculty of reproducing datasets, the

lack of in-depth explanations about the relationship between AI techniques and these ITS data, the need to

deepen these techniques of AI, and the need for more research about software engineering to ITS. This work

contributes to providing opportunities to this research area that can help the digital transformation of Computer

Science Education.

1 INTRODUCTION

The challenges of teaching Computer Science for stu-

dents and teachers can be perceived when analyz-

ing the various disciplines in the area. The difﬁ-

culties that students experience during the teaching-

learning process in the disciplines of algorithms and

data structures are a signiﬁcant concern for educators

around the world (Silva et al., 2019). About teaching

Software Engineering, the study (Ouhbi and Pombo,

2020) interviewed educators in the area and found dif-

ﬁculties in involving students in Software Engineer-

ing courses and in developing practical activities for

students. These two examples highlight some prob-

lems that are related to Computer Science Education.

Intelligent Tutoring System (ITS) is an educa-

tional software aimed at adapting teaching to the stu-

dent proﬁle (Alkhatlan and Kalita, 2018). Using Ar-

tiﬁcial Intelligence (AI), these systems can contribute

to the automation of stages of the teaching-learning

process in this area. For example, supporting a stu-

dent in solving problems involving practice and ab-

stract reasoning can be made possible by these sys-

https://orcid.org/0000-0003-2866-3431

https://orcid.org/0000-0001-7051-7396

tems.

This scenario makes it relevant to have a broad

view of ITS’s for Computer Science Education. How-

ever, we cannot ﬁnd many studies that review the

literature on ITS’s for Computer Science Education.

Moreover, the few studies found do not consider the

use of AI in this research area.

This literature review analyses the current sce-

nario of the development and use of ITS in Com-

puter Science Education, focusing on AI-based tech-

nics. This review contributes to the reﬂection on the

problems that hinder the advancement of this area of

research and the availability of these ITS’s for stu-

dents and teachers. The main contributions are:

• An overview of the development and use of ITS’s

in Computer Science Education focusing on the

use of AI-based techniques in this area;

• Discuss issues that hinder speed, alignment, and

collaboration in research on ITS’s for Computer

Science Education;

• A Discussion of the need to go further in the in-

vestigation of AI in these studies due to the lack of

clarity about the relationship between data models

and AI algorithms and the unavailability of public

datasets to improve AI adoption and results;

338

Francisco, R. and Silva, F.

Intelligent Tutoring System for Computer Science Education and the Use of Artiﬁcial Intelligence: A Literature Review.

DOI: 10.5220/0011084400003182

In Proceedings of the 14th International Conference on Computer Supported Education (CSEDU 2022) - Volume 1, pages 338-345

ISBN: 978-989-758-562-3; ISSN: 2184-5026

• An indication of possible strategies to deal with

these problems to advance the area.

We organized this work as follows: Section 2 de-

scribes the background on ITS. Section 3 describes

the planning to this review. Section 4 describes the

current scenario of ITS’s in Computer Science Edu-

cation. Section 5 focus on the use of the AI in ITS’s

for computer science. Section 6 describes challenges

and opportunities. Finally, Section 7 describes some

concluding remarks.

2 BACKGROUND

This section addresses the foundations for under-

standing this literature review. We will present the

ITS main modules and related work.

2.1 Intelligent Tutoring System (ITS)

According to (Alkhatlan and Kalita, 2018), ITS’s

are software capable of improving, adapting, and au-

tomating teaching. These systems offer the student

interaction with the content adapted to their proﬁle to

improve the learning experience.

An ITS usually has the Modules: Tutor, Student,

Expert Knowledge, and User Interface. We consider

these module names to standardize and facilitate un-

derstanding in this work. The Tutor Module is re-

sponsible for making pedagogical decisions and ex-

ecuting pedagogical instruction. This module needs

algorithms to deﬁne the ITS intervention threshold

in the study process, operationalize this intervention,

and recommend content. The Student Module needs

to capture, update and retrieve the student’s proﬁle.

The Expert Knowledge Module needs to represent

and manipulate the learning content. The Student and

Expert Knowledge Modules provide input for the Tu-

tor Module to perform its tasks. The User Interface

Module enables students to interact with the content

and with the behavior of an ITS.

ITS literature presents different nomenclatures

when describing the components of an ITS. A com-

mon term is model instead of module. The Expert

Knowledge Module (EKM) sometimes is called the

domain model.

2.2 Related Work

We searched on surveys and reviews in order to ver-

ify how these works approached the topic of ITS’s for

Computer Science Education. We found six surveys

about ITS, and only two of which (Crow et al., 2018;

Nesbit et al., 2015) addressed the domain of Com-

puter Science Education.

Surveys (Almasri et al., 2019; Alkhatlan and

Kalita, 2018; Mousavinasab et al., 2021; Kulik and

Fletcher, 2016) address general aspects of ITS’s.

They do not emphasize ITS’s for Computer Science

Education and therefore are unrelated works.

The work of (Crow et al., 2018) presented a Sys-

tematic Review on ITS’s for the computer program-

ming domain. This analysis showed the distinction

between these ITS’s concerning the topics covered

and the resources offered to students.

The study (Nesbit et al., 2015) reported on an on-

going Survey on ITS’s for Computer Science Edu-

cation and Software Engineering Education. Com-

puter programming was the discipline with the high-

est number of ITS proposals. They noticed a gradual

increase in publications on this subject from 1975 to

2014. This survey aimed to analyze aspects related to

the student model, content mastery, and the incorpo-

ration of technological resources ranging from games

to the analysis of the student’s affective state.

Despite the relevance of these surveys, we noticed

gaps in the research. The most recent survey (Crow

et al., 2018) only addressed the domain of computer

programming, and the survey (Nesbit et al., 2015)

mentioned that the most signiﬁcant number of ITS’s

among those analyzed are also for the domain of pro-

gramming. Both analyzed ITS’s features, but neither

emphasized the AI techniques used in these features.

This present survey investigates this gap by analyzing

research on ITS’s for Computer Science Education,

focusing on the use of AI techniques.

3 REVIEW PLANNING

This Literature Review aims to analyze the research

scenario on ITS’s for Computer Science Education.

The following research questions (RQ) were the basis

for this work.

• RQ1 - What is the current scenario about the de-

velopment and the use of ITS’s in Computer Sci-

ence Education?

• RQ2 - How is the use of AI in ITS’s for Computer

Science Education?

Some available researches propose educational

technologies whose didactic content to involved is

generic, making it possible to diversify the contents.

However, this research is interested in speciﬁc ITS’s

for Computer Science Education. These ITS’s have a

design planned concerning the content.

Intelligent Tutoring System for Computer Science Education and the Use of Artiﬁcial Intelligence: A Literature Review

339

This review relied on a Google Scholar search. We

chose Google Scholar because it indexes the leading

databases of scientiﬁc papers. Table 1 presents each

search string and its respective number of hits.

Table 1: Search strings and Number of Hits.

Search String Nº Hits

String"intelligent tutoring systems" computer science 59600

"intelligent tutoring systems" "computer science" 31600

"intelligent tutoring system" computer science 30600

"intelligent tutoring system" "computer science" 17200

"intelligent tutoring systems" "software engineering" 7270

"intelligent tutoring system" "software engineering" 4350

"intelligent tutoring systems" "software maintenance" 321

"intelligent tutoring system" "software maintenance" 170

We deﬁned these search strings due to the interest

in analyzing the context of ITS’s for Computer Sci-

ence Education in a conceptual and practical perspec-

tive. The term ITS has been combined with computer

science to reﬂect the conceptual perspective and has

been combined with software engineering and soft-

ware maintenance to reﬂect the practical view.

The software engineering and software mainte-

nance contents are potentially subsets of the computer

science content set, which makes this set of search

strings a reinforcement for the search as a whole. The

interest in analyzing works with a practical perspec-

tive related to this area also occurred due to the im-

portance of these valuable skills for professionals in

the industry.

We chose to analyze the ﬁrst ten resulting pages,

as the results are ordered by relevance, and these

searches brought a vast number of results. All the

search results had the title, and abstract analyzed to

verify their use in this work. This process generated

an initial set of 40 papers, and after further reading,

we reduced this initial set to the 26 articles used in

this literature review.

4 CURRENT SCENARIO OF ITS’S

IN COMPUTER SCIENCE

EDUCATION

To understand the current scenario of use and de-

velopment of ITS’s in Computer Science Education

(RQ1), we made the Table 2. Table 2 lists the

works found that fall under tutor systems and ITS’s

with application domains aligned with the Associa-

tion for Computing Machinery (ACM) (ACM Com-

puting Curricula Task Force, 2013) curriculum. We

also describe whether the work presents and details

the three main modules of the ITS architecture.

According to the works found, we noticed the ex-

istence of two categories of works. The ﬁrst category

clariﬁes that the work presents an ITS according to its

conceptual deﬁnition. However, the second category

does not explain its ITS, so it only describes a tutor

system.

Among the works, some of them presented a ITS’s

for Computer Science Education (Jeremic et al., 2009;

Jeremic et al., 2012; Carter and Blank, 2013; Hars-

ley et al., 2016; Abd Rahman et al., 2016; Verdú

et al., 2017; Price et al., 2017; Hooshyar et al., 2018;

Figueiredo and García-Peñalvo, 2020; Galafassi et al.,

2020; Alshaikh et al., 2021). Modules found in these

works are aligned with the conceptual deﬁnition of

ITS modules. They present dynamics based on algo-

rithms and data representations that enable the execu-

tion of the ITS’s functionalities.

As described in Table 2, several computer science

related contents were addressed in the ITS’s. Among

these contents, it is possible to exemplify: Computer

Network Design (Verdú et al., 2017) and Program-

ming (Figueiredo and García-Peñalvo, 2020).

We found ﬁve works (Alshaikh et al., 2021;

Figueiredo and García-Peñalvo, 2020; Hooshyar

et al., 2018; Price et al., 2017; Carter and Blank,

2013) whose application domains of ITS’s are aligned

to the ACM Body of Knowledge described as Soft-

ware Development Fundamentals.

Within these ﬁve works within the domain of

Software Development Fundamentals, we found that

three systems (Figueiredo and García-Peñalvo, 2020;

Hooshyar et al., 2018; Price et al., 2017) that ad-

dress the ﬁeld of computer programming. The ITS

presented by the study by (Figueiredo and García-

Peñalvo, 2020), which was built as a teacher support

tool, can detect the student’s situation, needs, skills,

and learning style from a predictive model to con-

tribute to the work of the teacher who needs to decide

the next steps of instruction. The solution-based ITS

proposed by (Hooshyar et al., 2018) seeks to improve

students’ programming problem-solving strategy. It

offers solutions that help students with the automatic

generation of ﬂowcharts that represent the algorithms

related to the exercises provided by the tool and sup-

port the navigation of topics and activities. The study

on the programming environment for newbies iSnap

(Price et al., 2017) presented the incorporation of the

idea of offering tips during task resolution, from the

ITS, in its proposal.

The other studies found for Software Develop-

ment Fundamentals address Software Understanding

(Alshaikh et al., 2021), and Software Debugging

(Carter and Blank, 2013). The work by (Alshaikh

et al., 2021) presented a Socratic ITS, its author-

CSEDU 2022 - 14th International Conference on Computer Supported Education

340

Table 2: Papers on ITS for Computer Science Education.

Reference ACM Body of Knowledge ACM Discipline

Detail All

ITS Modules?

ITS Module

Tutor? Student?

Expert

Knowledge?

(Nakhal and Bashhar, 2017) Algorithms and Complexity

Basic Automata, Computability and

Complexity; Advanced Automata

Theory and Computability

No Yes No No

(Harsley et al., 2016) Algorithms and Complexity

Fundamental Data Structures and

Algorithms

No No Yes No

(Albatish et al., 2018) Architecture and Organization Digital Logic and Digital Systems No Yes No No

(Galafassi et al., 2020) Discrete Structures Proof Techniques No No Yes No

(AbuEl-Reesh and Abu-Naser, 2018)

Information Assurance and

Security

Cryptography No Yes No No

(Elreesh and Abu-Naser, 2019)

Information Assurance and

Security

Cryptography No Yes No No

(Mahdi et al., 2016)

Information Assurance and

Security

Threats and Attacks; Cryptography No No No No

(Al-Hanjori et al., 2017)

Networking and

Communication

Introduction; Local Area Networks No No No No

(Alshawwa et al., 2019)

Networking and

Communication

Introduction; Local Area Networks No Yes No No

(Verdú et al., 2017)

Networking and

Communication

Introduction; Local Area Networks Yes Yes Yes Yes

(Marouf and Abu-Naser, 2019) Operating Systems Overview of Operating Systems No Yes No No

(Hasanein and Naser, 2017)

Parallel and Distributed

Computing

Cloud Computing No Yes No No

(Oberhauser, 2017)

Software Development

Fundamentals

Development Methods No No No No

(Haddad and Naser, 2017)

Software Development

Fundamentals

Development Methods No Yes No No

(Carter and Blank, 2013)

Software Development

Fundamentals

Development Methods Yes Yes Yes Yes

(Paaßen et al., 2016)

Software Development

Fundamentals

Fundamental Programming

Concepts

No No No No

(Al-Bastami and Naser, 2017)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No No

(Price et al., 2017)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No No

(Mosa et al., 2018)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No No

(Hooshyar et al., 2018)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No Yes

(Al-Shawwa et al., 2019)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No No

(Figueiredo and García-Peñalvo, 2020)

Software Development

Fundamentals

Fundamental Programming

Concepts

No No Yes No

(Alshaikh et al., 2021)

Software Development

Fundamentals

Fundamental Programming

Concepts

No Yes No Yes

(Jeremic et al., 2009) Software Engineering Software Design Yes Yes Yes Yes

(Jeremic et al., 2012) Software Engineering Software Design Yes Yes Yes Yes

(Abd Rahman et al., 2016) Software Engineering Software Project Management Yes Yes Yes Yes

ing tool and its mechanism of generation and under-

standing dialogues for the domain of software com-

prehension. The ITS for software debugging (Carter

and Blank, 2013) supports individually the student in

problem solving.

For the Software Engineering domain, the ITS’s

(Abd Rahman et al., 2016; Jeremic et al., 2012;

Jeremic et al., 2009) were found. The ITS ABITS-

FPM (Abd Rahman et al., 2016) was proposed for

teaching and learning Metrics for Function Points.

This ITS provides visualization, immediate feedback,

recommendation, interactive help, and guided help.

The paper presented: a student model composed of

student personal facts, learning levels, presentation

styles, and assessment results; and a domain model

aligned with knowledge about Metrics for Function

Points, which includes theoretical content, practice

questions, question tips, and supporting information

related to student practical performance.

The ITS DEPTHS (Jeremic et al., 2012; Jeremic

et al., 2009) was proposed for the topic of software

design patterns. The study by (Jeremic et al., 2009)

presented the use of a dependency graph to model

the domain of software design patterns, proposed a

strategy for the tutor module that involves facts, rules,

queries, and the production of concept plans and plans

of class and the accomplishment of tests to evaluate

the student that collects data of difﬁculty, time and

result. However, the (Jeremic et al., 2012) study em-

phasized the student model at ITS DEPTHS with the

use of personal data, static and dynamic performance

data, and teaching histories and their updating.

However, we found only one work for each of

the other domains, i.e., Algorithms and Complexity

(Harsley et al., 2016), Discrete Structures (Galafassi

et al., 2020) and Networking and Communication

(Verdú et al., 2017).

The Collaborative ITS Collab-ChiQat Tutor was

proposed for the teaching-learning of algorithms, and

basic data structures (Harsley et al., 2016). This ITS,

in addition to including a very intuitive User Interface

(UI) that integrates programming and visualization of

Intelligent Tutoring System for Computer Science Education and the Use of Artiﬁcial Intelligence: A Literature Review

341

data structures, presented a student model that con-

siders the individual and collaborative behavior of the

student during the use of the system.

The study on the ITS EvoLogic (Galafassi et al.,

2020) addressed the teaching-learning of Natural De-

duction in Propositional Logic. They presented model

tracing as a resource capable of tracking the individ-

ual steps of each student to provide real-time feed-

back and the student model, which represents these

steps performed by students, categorizing the quality

of their efforts and their line of reasoning during the

task resolution.

The INTUITEL (Verdú et al., 2017) approach,

which includes an ITS adaptable to the Learning

Management System (LMS) and performs content

recommendation, was applied to a Computer Net-

work Design course through integration with Moo-

dle. This approach offers non-intrusive recommen-

dations and feedback on the best learning path con-

sidering proﬁle, progress, context, pedagogical strate-

gies, and environmental inﬂuences. Teachers need to

model the learning process in INTUITEL concern-

ing their teaching materials and strategies, and IN-

TUITEL handles this modeling in an ontology-based

approach. Although the system has been applied in

the Computer Network Design domain and integrated

into the LMS Moodle, it can be used in other courses

and integrated into other LMS’s.

We found in the literature on ITS’s for Computer

Science Education that describe Tutor Systems that,

despite claiming that they are ITS’s, approach the

characteristics of ITS’s in a very simpliﬁed way, e.g.

without presenting computational techniques such as

Artiﬁcial Intelligence algorithms or structures for rep-

resentation of models. The papers (Marouf and Abu-

Naser, 2019; Elreesh and Abu-Naser, 2019; Mosa

et al., 2018; AbuEl-Reesh and Abu-Naser, 2018; Al-

batish et al., 2018; Hasanein and Naser, 2017; Haddad

and Naser, 2017; Al-Hanjori et al., 2017; Nakhal and

Bashhar, 2017; Mahdi et al., 2016), which were based

on the authoring tool Intelligent Tutoring System

Builder (ITSB), and work (Alshawwa et al., 2019; Al-

Shawwa et al., 2019; Paaßen et al., 2016; Oberhauser,

2017) ﬁt into this perspective.

In this sense, for this Tutor Systems category, only

the works (Oberhauser, 2017; Paaßen et al., 2016)

will be presented as they present future possibilities to

contribute to the construction of ITS’s for Computer

Science Education.

A source code recommendation, navigation, and

3D visualization approach (Oberhauser, 2017) was

proposed to contribute to the understanding of soft-

ware in a perspective of exploratory, analytical, and

descriptive cognitive processes. This Tutor System

has a recommendation service based on a theoretical

model of program comprehension and information re-

lated to the MethodRank metric for source code, ﬁlter

processing, calculations of distances, and points of in-

terest in the source code.

The strategy presented by the study by (Paaßen

et al., 2016), which addressed the tracking of pro-

gram execution, can be used in ITS’s for the teaching-

learning of computer programming. In this strategy,

the execution traces of the programs are captured and

compared using the Edit Distance algorithm to obtain

information about the functionality that the program

implements, regardless of syntactic differences.

5 AI IN ITS’S FOR COMPUTER

SCIENCE

Table 3 aims to help understand the scenario of the

use of AI in ITS’s for Computer Science Education

(RQ2). Table 3 presents the subset of works from

Table 2 that addressed AI in the ITS’s. Information

about the AI technique used, the existence of the al-

gorithm description, the availability or possibility of

reproducing the datasets, and the availability of the

ITS for use are presented.

Agent-based architecture was an AI technique

used in the (Hasanein and Naser, 2017; Abd Rahman

et al., 2016) studies. Despite using this type of ar-

chitecture, these studies did not mention which algo-

rithms were used in their agents.

Multiagent System and Bayesian Network were

other AI techniques found in one study. The ITS

based on solutions for the teaching-learning of com-

puter programming (Hooshyar et al., 2018) used a

Multiagent System for the automatic generation of

ﬂowcharts from an approach that works with the text

of the problem speciﬁcation and a Bayesian Network

to model the programming domain content and pre-

requisites.

The Case-based reasoning (CBR) was used in an

ITS for software debugging (Carter and Blank, 2013).

The perception of the relationship between debugging

and CBR motivated the authors to propose CBR in the

design of this ITS. The proposal includes the CBR

cycle and uses representations of the software and er-

rors, e.g., the generated abstract syntax tree, compila-

tion, execution, and static analysis data.

Fuzzy was used by the (Jeremic et al., 2012;

Jeremic et al., 2009) studies, which are about the ITS

DEPTHS. This ITS works with the updating of the

student model through Fuzzy rules (Jeremic et al.,

2012), which involves production rules and Fuzzy

logic, and with Fuzzy sets and theories of certainty,

CSEDU 2022 - 14th International Conference on Computer Supported Education

342

Table 3: Papers that addressed AI.

Reference AI Based Technique Algorithm

Dataset

Available

Dataset

Reproducible

ITS

Available

(Hasanein and Naser, 2017) Agent-based architecture No No No No

(Abd Rahman et al., 2016) Agent-based architecture No No No No

(Hooshyar et al., 2018)

Bayesian networks;

Multi-agent system

Superﬁcially

presented

No No No

(Carter and Blank, 2013) Case-based reasoning No No Yes No

(Jeremic et al., 2009) Fuzzy sets No No No No

(Jeremic et al., 2012) Fuzzy rules Yes No Yes No

(Galafassi et al., 2020) Genetic Algorithm Yes No Yes No

(Figueiredo and García-Peñalvo, 2020) Neural network Yes No Yes No

(Verdú et al., 2017) Ontology-based reasoning No No No Yes

factors to calculate the student’s knowledge status

(Jeremic et al., 2012; Jeremic et al., 2009).

Genetic Algorithm was used by ITS EvoLogic

(Galafassi et al., 2020). This study proposes model

tracing as a resource for the expert agent, which uses

an adapted GA to solve Natural Deduction problems

in the Propositional Logic to accompany students in

real-time. Neural Network was used in the ITS for the

domain of introduction to programming (Figueiredo

and García-Peñalvo, 2020). The neural network-

based model predict student failure.

The Ontology-based reasoning approach was used

by the INTUITEL (Verdú et al., 2017) project. This

project, which includes an ITS adaptable to the

LMS and performs content recommendation, uses

ontology-based reasoning from various ontologies or-

ganized in layers. This reasoning is made possible by

the INTUITEL Reasoning Engine and the Learning

Progress Model (LPM). The LPM deduces the stu-

dent’s learning state so that the INTUITEL Reason-

ing Engine can use it to recommend the next learning

steps.

6 CHALLENGES AND

OPPORTUNITIES

As a result of this study, this section presents some re-

search challenges and opportunities identiﬁed in this

area.

6.1 Low use of AI Techniques

There is low usage of AI techniques in ITS’s for Com-

puter Science Education. Among the 26 articles in-

cluded in this review, only nine addressed some AI

techniques, equivalent to 34.6 %.

In addition to the few articles that addressed AI

techniques, we noticed that 45.5% of the articles that

used AI did not even present details of the algorithms.

The works (Hasanein and Naser, 2017; Abd Rahman

et al., 2016), which used Agent-based architecture,

did not even mention which algorithms were used in

their agents.

AI techniques are impacting several areas and we

can clearly state that there are many different oppor-

tunities to explore its use in ITS for Computer Sci-

ence Education, then improving and strengthening

this area.

6.2 ITS’s not Available Publicly

Computer Science Education has geographic particu-

larities, which requires investigating the effects these

ITS and their AI-based functionalities bring to teach-

ing in different contexts. Only one of the AI-based

ITS’s was publicly available for use (Verdú et al.,

2017). This fact hampers the popularization of ITS’s

in real educational scenarios, partnerships between re-

searchers, and research advances.

In this sense, it is welcome that the ITS’s could

be available to other researchers and teachers, which

could improve the research and development of this

area. At least, the research community could have at

least this goal for new systems and techniques.

6.3 Lack of Public Datasets

None of these studies with the ITS’s that addressed

AI presented the datasets used in the AI algorithms.

Therefore, researchers need to reconstruct scenarios

to collect data that make it possible to analyze the re-

sult of new AI techniques or even improvements in

existing methods.

This lack of public data hinders the comparison

between different AI techniques. A welcomed op-

portunity here is to capture and make public datasets

obtained from real scenarios of the use of ITS’s for

Computer Science Education.

Intelligent Tutoring System for Computer Science Education and the Use of Artiﬁcial Intelligence: A Literature Review

343

6.4 Difﬁculty to Reproduce Datasets

Another challenge is to reproduce datasets using

ITS’s presented in the literature. In the set of stud-

ies on ITS’s for this domain using AI, only 44.4% of

these works described data or text that could help in

this reproduction. In addition, there is the difﬁculty

of accessing educational scenarios and modifying the

teaching-learning process to carry out this data col-

lection from the use of ITS’s by students.

Reproducing scenarios could provide a systematic

opportunity to compare different ITS’s, indicating the

right direction to improvements and ﬁnding learned

lessons that the following research should avoid.

6.5 Low Coverage of Topics

The surveyed ITS’s covers only 44.4% (8 out of 18) of

the ACM Body of Knowledge several topics. Further-

more, the group of disciplines ACM Body of Knowl-

edge named Software Development Fundamentals oc-

cupies 30.7% of these surveys.

This fact brings the opportunity to expand the de-

sign, development, and deployment into other groups

of disciplines in the Computer Science area.

7 CONCLUDING REMARKS

In this work, we conducted a literature review about

the AI-based techniques used in ITS for Computer

Science Education, considering the undergraduate

courses. Our work contributes to this area by pre-

senting an analysis of this research area with the AI

perspective. This research found 26 different articles

in the review scope.

Our ﬁndings of the current scenario of develop-

ment and use of ITS’s for Computer Science Educa-

tion show that these systems’ set of disciplines are

small. Only 19.2 % of the analyzed ITS’s detail the

three main modules of its architecture. Only 26.9 %

of the papers present the EKM, which represents and

manages the educational content. The review consid-

ered not only ITS’s but also tutor systems that follow

a different concept compared to the ITS.

We noticed some AI methods in these studies,

such as Genetic Algorithms, Bayesian Networks, and

Neural Networks. In general, the works lack descrip-

tions of the used techniques in ITS’s needing further

explanation. None of the articles available the dataset.

The papers presented little information about AI algo-

rithms. We found only one ITS that used AI and was

publicly available. There are several opportunities for

the use of AI in ITS.

Our assessment of this research area brought a set

of challenges and opportunities. There is room to im-

prove the use of the AI techniques on ITS’s improving

presented and using these techniques in researching

new systems.

Another opportunity that the research community

can explore is to publish ITS’s, make them pub-

lic, and foster their use in different educational con-

texts. This usage can contribute to the lack of pub-

lic datasets that could promote research on new AI

techniques and create reproducible experiments using

public datasets. These opportunities could also con-

tribute to improving experimental research in this re-

search area.

Finally, a clear opportunity in this area is to ex-

pand the coverage of the ITS to other topics of Com-

puter Science Education. This expansion would im-

prove the teaching-learning process in several areas of

knowledge that encompass computer science, helping

the digital transformation of Computer Science Edu-

cation.

ACKNOWLEDGEMENTS

This work is inside UFU-CAPES.Print Program. This

study was ﬁnanced in part by the Coordenação de

Aperfeiçoamento de Pessoal de Nível Superior –

Brasil (CAPES) – Finance Code 001. This research

also received the support from PROPP/UFU.

REFERENCES

Abd Rahman, A., Abdullah, M., and Alias, S. H. (2016).

The architecture of agent-based intelligent tutoring

system for the learning of software engineering func-

tion point metrics. In 2016 2nd International Sym-

posium on Agent, Multi-Agent Systems and Robotics

(ISAMSR), pages 139–144. IEEE.

AbuEl-Reesh, J. Y. and Abu-Naser, S. S. (2018). An Intelli-

gent Tutoring System for Learning Classical Cryptog-

raphy Algorithms (CCAITS). 2(2):11.

ACM Computing Curricula Task Force, editor (2013).

Computer Science Curricula 2013: Curriculum

Guidelines for Undergraduate Degree Programs in

Computer Science. ACM, Inc.

Al-Bastami, B. G. and Naser, S. A. (2017). Design and

Development of an Intelligent Tutoring System for C#

Language. undeﬁned.

Al-Hanjori, M. M., Shaath, M. Z., and Abu-Naser, S. S.

(2017). Learning computer networks using intelligent

tutoring system.

Al-Shawwa, M., Alshawwa, I. A., and Abu-Naser, S. S.

(2019). An Intelligent Tutoring System for Learning

Java. 3(1):6.

CSEDU 2022 - 14th International Conference on Computer Supported Education

344

Albatish, I., Mosa, M. J., and Abu-Naser, S. S. (2018).

ARDUINO Tutor: An Intelligent Tutoring System for

Training on ARDUINO. 2(1):10.

Alkhatlan, A. and Kalita, J. (2018). Intelligent tutoring sys-

tems: A comprehensive historical survey with recent

developments. arXiv preprint arXiv:1812.09628.

Almasri, A., Ahmed, A., Almasri, N., Abu Sultan, Y. S.,

Mahmoud, A. Y., Zaqout, I. S., Akkila, A. N., and

Abu-Naser, S. S. (2019). Intelligent Tutoring Sys-

tems Survey for the Period 2000- 2018. Accepted:

5/28/2019 Publisher: IJARW.

Alshaikh, Z., Tamang, L., and Rus, V. (2021). Experiments

with Auto-generated Socratic Dialogue for Source

Code Understanding:. In Proceedings of the 13th In-

ternational Conference on Computer Supported Edu-

cation, pages 35–44, Online Streaming, — Select a

Country —. SCITEPRESS - Science and Technology

Publications.

Alshawwa, I. A., Al-Shawwa, M., and Abu-Naser, S. S.

(2019). An Intelligent Tutoring System for Learning

Computer Network CCNA. 3(2):9.

Carter, E. and Blank, G. D. (2013). An intelligent tutor-

ing system to teach debugging. In International Con-

ference on Artiﬁcial Intelligence in Education, pages

872–875. Springer.

Crow, T., Luxton-Reilly, A., and Wuensche, B. (2018).

Intelligent tutoring systems for programming educa-

tion: a systematic review. In Proceedings of the

20th Australasian Computing Education Conference,

pages 53–62.

Elreesh, J. Y. A. and Abu-Naser, S. S. (2019). Cloud Net-

work Security Based on Biometrics Cryptography In-

telligent Tutoring System. 3(3):34.

Figueiredo, J. and García-Peñalvo, F. J. (2020). Intelligent

tutoring systems approach to introductory program-

ming courses. In Eighth International Conference on

Technological Ecosystems for Enhancing Multicultur-

ality, pages 34–39.

Galafassi, C., Galafassi, F., Reategui, E., and Vicari, R.

(2020). Evologic: Sistema tutor inteligente para en-

sino de lógica. In Anais do XLVII Seminário Integrado

de Software e Hardware, pages 222–233. SBC.

Haddad, I. A. E. and Naser, S. A. (2017). ADO-Tutor: In-

telligent Tutoring System for leaning ADO.NET. un-

deﬁned.

Harsley, R., Di Eugenio, B., Green, N., Fossati, D., and

Acharya, S. (2016). Integrating support for collabora-

tion in a computer science intelligent tutoring system.

In International conference on intelligent tutoring sys-

tems, pages 227–233. Springer.

Hasanein, H. A. A. and Naser, S. S. A. (2017). An intelli-

gent tutoring system for cloud computing. page 5.

Hooshyar, D., Ahmad, R. B., Youseﬁ, M., Fathi, M., Horng,

S.-J., and Lim, H. (2018). Sits: A solution-based in-

telligent tutoring system for students’ acquisition of

problem-solving skills in computer programming. In-

novations in Education and Teaching International,

55(3):325–335.

Jeremic, Z., Jovanovic, J., and Gasevic, D. (2009). Evaluat-

ing an intelligent tutoring system for design patterns:

The depths experience. Journal of Educational Tech-

nology & Society, 12(2):111.

Jeremic, Z., Jovanovi

c, J., and Gaševi

c, D. (2012). Stu-

dent modeling and assessment in intelligent tutoring

of software patterns. Expert Systems with Applica-

tions, 39(1):210–222.

Kulik, J. A. and Fletcher, J. (2016). Effectiveness of intelli-

gent tutoring systems: a meta-analytic review. Review

of educational research, 86(1):42–78.

Mahdi, A. O., Alhabbash, M. I., and Naser, S. S. A. (2016).

An intelligent tutoring system for teaching advanced

topics in information security. page 9.

Marouf, A. M. and Abu-Naser, S. S. (2019). Intelligent

tutoring system for teaching computer science i in al-

azhar university, gaza. International Journal of Aca-

demic and Applied Research (IJAAR), 3(3):31–53.

Mosa, M. J., Albatish, I., and Abu-Naser, S. S. (2018).

ASP.NET-Tutor: Intelligent Tutoring System for

leaning ASP.NET. 2(2):8.

Mousavinasab, E., Zarifsanaiey, N., R. Niakan Kalhori,

S., Rakhshan, M., Keikha, L., and Ghazi Saeedi,

M. (2021). Intelligent tutoring systems: a system-

atic review of characteristics, applications, and eval-

uation methods. Interactive Learning Environments,

29(1):142–163.

Nakhal, M. and Bashhar, B. (2017). Adaptive ITS for

Learning Computer Theory. European Academic Re-

search, 4(10):8770–8782.

Nesbit, J., Liu, L., Liu, Q., and Adesope, O. O. (2015).

Work in Progress: Intelligent Tutoring Systems in

Computer Science and Software Engineering Educa-

tion.

Oberhauser, R. (2017). Visitr: 3d visualization for code

visitation trail recommendations. International Jour-

nal on Advances in Software Volume 10, Number 1 &

2, 2017.

Ouhbi, S. and Pombo, N. (2020). Software Engineer-

ing Education: Challenges and Perspectives. In

2020 IEEE Global Engineering Education Confer-

ence (EDUCON), pages 202–209. ISSN: 2165-9567.

Paaßen, B., Jensen, J., and Hammer, B. (2016). Execution

traces as a powerful data representation for intelligent

tutoring systems for programming. International Ed-

ucational Data Mining Society.

Price, T. W., Dong, Y., and Lipovac, D. (2017). isnap: to-

wards intelligent tutoring in novice programming en-

vironments. In Proceedings of the 2017 ACM SIGCSE

Technical Symposium on computer science education,

pages 483–488.

Silva, D. B., Aguiar, R. d. L., Dvconlo, D. S., and Silla,

C. N. (2019). Recent Studies About Teaching Algo-

rithms (CS1) and Data Structures (CS2) for Computer

Science Students. In 2019 IEEE Frontiers in Educa-

tion Conference (FIE), pages 1–8. ISSN: 2377-634X.

Verdú, E., Regueras, L. M., Gal, E., de Castro, J. P., Verdú,

M. J., and Kohen-Vacs, D. (2017). Integration of an

intelligent tutoring system in a course of computer

network design. Educational Technology Research

and Development, 65(3):653–677.

Intelligent Tutoring System for Computer Science Education and the Use of Artiﬁcial Intelligence: A Literature Review

345