Revue internationale sur le numérique en éducation et communication
revue-mediations.teluq.ca No 16, Autumn 2023
30
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
International Journal of Digital Education and Communication
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
1
AReviewoftheLiteraturefrom1970to
2022ontheRolesofTeachersand
ArtificialIntelligenceintheFieldofAIin
Education
Unerecensiondesécritsde1970à2022surlesrôlesde
l’enseignantetdel’intelligenceartificielledansledomainede
l’IAenéducation
Unarevisióndelosescritosde1970a2022sobrelospapeles
delprofesorylainteligenciaartificialenelcampodelaIAen
educación
https://doi.org/10.52358/mm.vi16.304
Alexandre Lepage, doctoral student in education
Université de Montréal, Canada
alexandre.lepage.2@umontreal.ca
Normand Roy, Professor
Université de Montréal, Canada
normand.roy@umontreal.ca
ABSTRACT
This article reviews the literature on the role of artificial intelligence (AI) and what teachershave
envisioned in the field of artificial intelligence in education (AIED) since 1970. Forty-eight
documents, most of them theoretical, were analyzed to identify what roles are given to AI in
relation to learners, teachers, knowledge and the classroom as a whole (i.e. supporting
motivation or providing personalized feedback). Quotes discussing teachers’ role toward these
components of learning situations were also analyzed (i.e. orchestrating interactions or
evaluating learners). The results show considerable overlap between teachers’ role and what
AI is being developed to achieve in the field of AIED. Even if impossible in a predictable future,
the ambition of research in the field seems to be to automate a growing number of teachers’
revue-mediations.teluq.ca No 16, Autumn 2023
31
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
2
tasks. In the meantime, the role of teachers appears to be a dead angle in the field of AIED.
The discussion proposes to reuse Faerber’s ICT pyramid (2003), based on Houssaye’s
didactic triangle (1988/2015), to better study the role of AI in education in relation to those of
teachers and learners.
Keywords: artificial intelligence, education, teacher, artificial intelligence in education, ICT
RÉSUMÉ
Cet article présente une recension des écrits sur la façon dont les rôles de l’enseignant et de
l’intelligence artificielle (IA) sont abordés dans le domaine de l’intelligence artificielle en
éducation (IAED) depuis 1970. Quarante-sept documents, théoriques pour la plupart, ont été
analysés à partir des passages relatifs aux tâches confiées à une IA en lien avec les
apprenants, les enseignants, les savoirs ou la classe (p. ex. le soutien à la motivation ou la
rétroaction personnalisée). Les passages qui discutent du rôle de l’enseignant en lien avec
ces différentes composantes ont aussi été analysés (p. ex. le pilotage des interactions en
classe ou l’évaluation des apprenants). Les principaux résultats montrent que les
développements dans le domaine de l’IAED couvrent un large spectre des attributions de
l’enseignant et que, même si inatteignable sur un horizon prévisible, l’ambition scientifique du
domaine semble être d’automatiser de plus en plus de tâches de l’enseignant. Il ressort que
le rôle de l’enseignant est très peu discuté dans le domaine, et encore moins les interactions
attendues entre enseignants et IA. La discussion propose de réemployer le tétraèdre des TIC
en éducation de Faerber (2003), lui-même appuyé sur le triangle didactique de Houssaye
(1988), pour conceptualiser le rôle de l’IA en éducation en interaction avec ceux de
l’enseignant et de l’apprenant.
Mots-clés : intelligence artificielle, éducation, enseignant, intelligence artificielle en éducation,
TIC
RESUMEN
Este estudio presenta una revisión de la literatura sobre cómo se han abordado los papeles
del docente y de la inteligencia artificial (IA) en el campo de la inteligencia artificial en
educación (AIED) desde 1970. Se han analizado cuarenta y ocho artículos, en su mayoría
teóricos, a partir de los pasajes relacionados con las tareas encomendadas a una IA en
relación con los alumnos, los profesores, el conocimiento o la clase (por ejemplo, apoyo a la
motivación o retroalimentación personalizada). También se han analizado los pasajes que
discuten el papel del profesor en relación con estos diferentes componentes (por ejemplo, la
gestión de las interacciones en el aula o la evaluación de los alumnos). Los principales
resultados muestran que los desarrollos en el campo de la IAED cubren un amplio espectro
de atribuciones docentes y que, aunque inalcanzable en un horizonte previsible, la ambición
científica en el ámbito parece ser automatizar cada vez más las funciones docentes. Parece
que el papel del docente es muy poco discutido en el campo, y menos todavía las
interacciones esperadas entre los docentes y la IA. La discusión propone reutilizar el tetraedro
de las TIC en la educación de Faerber (2003), basado en el triángulo didáctico de Houssaye
(1988/2015), para conceptualizar el papel de la IA en educación en interacción con el del
docente y el del alumno.
Palabras clave: inteligencia artificial, educación, docente, inteligencia artificial en educación,
TIC
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
1
AReviewoftheLiteraturefrom1970to
2022ontheRolesofTeachersand
ArtificialIntelligenceintheFieldofAIin
Education
Unerecensiondesécritsde1970à2022surlesrôlesde
l’enseignantetdel’intelligenceartificielledansledomainede
l’IAenéducation
Unarevisióndelosescritosde1970a2022sobrelospapeles
delprofesorylainteligenciaartificialenelcampodelaIAen
educación
https://doi.org/10.52358/mm.vi16.304
Alexandre Lepage, doctoral student in education
Université de Montréal, Canada
alexandre.lepage.2@umontreal.ca
Normand Roy, Professor
Université de Montréal, Canada
normand.roy@umontreal.ca
ABSTRACT
This article reviews the literature on the role of artificial intelligence (AI) and what teachershave
envisioned in the field of artificial intelligence in education (AIED) since 1970. Forty-eight
documents, most of them theoretical, were analyzed to identify what roles are given to AI in
relation to learners, teachers, knowledge and the classroom as a whole (i.e. supporting
motivation or providing personalized feedback). Quotes discussing teachers’ role toward these
components of learning situations were also analyzed (i.e. orchestrating interactions or
evaluating learners). The results show considerable overlap between teachers’ role and what
AI is being developed to achieve in the field of AIED. Even if impossible in a predictable future,
the ambition of research in the field seems to be to automate a growing number of teachers’
revue-mediations.teluq.ca No 16, Autumn 2023
32
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
3
Introduction
Artificial intelligence in education (AIED) is a field of research whose beginnings can be traced back to the
decade 1971-1980 (Self, 2016). Since then, a great deal of research has been carried out, leading to the
development of a variety of digital tools including intelligent tutorial systems, conversational agents or,
more recently, educational success dashboards. The field follows in the footsteps of computer-aided
instruction and intelligent computer-aided instruction (Robertson, 1976) whose aim was to enable
knowledge to be learned on a computer, with exercisers providing automatic feedback. According to
Wenger (1986), unlike these two fields, the AIED field aimed to develop systems capable of making
instructional decisions themselves, rather than applying pre-programmed decisions. In 1987, Romiszowski
described the AIED field as one in which applications could serve either the teacher or the student, in one
of three ways: as a tutor (computer-assisted learning), as a tool (use of expert systems) or as a learning
object (learning to program a system). More recently, Lameras and Arnab (2021) reviewed the literature
for the period 2008-2020 and identified that the AIED field could be broken down into five sub-fields:
(1) content preparation and delivery, (2) helping students apply knowledge, (3) engaging students in tasks,
(4) assessment and feedback, and (5) helping learners self-regulate.
Changes in the AIED field have been numerous since its inception, starting with the multiplication of data
sources available to personalize learning according to individuals (Bull and Kay, 2016). This
personalization, a central objective of research in the field (Dede et al., 1985), is now based on a greater
quantity of parameters determined by increasingly complex digital traces, and by predictive models
established by massive data from a multitude of learners or use cases. The increasingly porous boundary
between digital and physical spaces (Dillenbourg, 2016) also leads us to see the field of AIED as less and
less hermetic, since the uses of AI in education can be studied via complex teaching-learning situations in
the classroom, whether face-to-face or virtual, or via other disciplines such as assessment or instructional
design. What’s more, the techniques used to operate AIED systems have evolved. Romiszowski (1987)
closely associated AIED with expert systems, which he defined as follows: “An expert system should help
a novice, or partly experienced, problem-solver to match acknowledged experts in the particular domain
of problem solving that the system is designed to assist” (p. 96). Today, the field of AIED is marked by the
growing use of so-called connectionist AI (see Minsky, 1991), via machine learning, giving rise to new uses
such as the prediction of academic success or the deployment of high-performance conversational agents.
The role of the teacher has received little attention in the field of AIED (du Boulay, 2021), with research
focusing mainly on learner-knowledge interactions mediated by intelligent tutorial systems that can take
the form of problem-solving assistants, mentors, laboratory assistants or expert consultants (Sleeman and
Brown, 1982). Yet teachers are central to the process of pedagogical integration of digital technology, and
any transfer of new technologies within the classroom starts with them. More and more systems aim to
support interactions between teachers and learners (Timms, 2016), which calls for a definition of the
teacher’s role in relation to the use of AI. Like other digital technologies, the pedagogical integration of AI
in education is only possible if it is first adopted by teachers. Secondly, it is through techno-pedagogical
integration, now studied by specific frameworks such as the T-PACK model (Koehler and Mishra, 2009),
that teachers will determine what is done before, during and after the use of a digital technology, with or
without AI. He or she will also determine what is expected of the learner at each of these stages, and may
combine several software applications to achieve a broader pedagogical objective than was intended when
the technology was designed. The deployment of AIED systems in the classroom therefore necessarily
involves teachers. But what is expected of them with regard to AI systems? What is their role? Does this
software encroach on teachers’ tasks? Are they better or worse than teachers at certain tasks? What new
responsibilities do teachers have when using AI systems?
revue-mediations.teluq.ca No 16, Autumn 2023
33
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
4
The relevance of this article lies mainly in the fear, founded or not, that AI could replace teachers. According
to Renz and Vladova (2021), this fear of teacher replacement has slowed the progress of AI in education
compared to other fields. For many, teachers are needed to manage particular situations that cannot be
anticipated (Holmes et al., 2021), make pedagogical decisions in the interests of the student and not the
Edtech sector (Saltman, 2020) or simply to preserve human interactions (Renz and Vladova, 2021). For
those people, the role of AI in education should rather be to support teachers in their actions (e.g., Bulger,
2016; Marrhich et al., 2021). But the distinction between teacher replacement and teacher support is not
so clear-cut in the case of a technology whose ambition is to reproduce a part of human intelligence. As
Mubin et al. (2013) point out, the division of labour between educational robots and teachers needs to be
clarified by going beyond the sterile dichotomy pertaining to replacement. The aim of the research is to
analyze how the roles of teachers and AI have been described in the AIED field between 1970 and 2022.
Achieving this objective will eventually help to identify areas of interaction between teachers and AI, as
well as gaps that must be filled in terms of roles with learners.
Method
The method selected was that of a systematic review based on the steps suggested by Rhoades (2011):
identification of inclusion and exclusion criteria, scanning of titles and abstracts to exclude irrelevant
studies, addition of references deemed missing, detailed analysis of relevant studies, data extraction,
synthesis and conclusion. An inductive analysis (Corbin and Strauss, 2015) was performed by establishing
a code grid, stabilized after the first 10 papers, with MaxQDA software. Title and abstract scanning was
carried out by one of the authors, as was document coding. The code grid was adjusted and validated by
both authors.
Databases consulted
The following databases were searched on November 1
st
2022: Web of Science (71 results), ACM Digital
Library (6), Science Direct (8), Erudit (3), Academic Search Premier and Education Source (39), Taylor
and Francis (4), as well as 11 manual additions, which were deemed to be missing (e.g., via references to
documents consulted or suggested during peer review). The search yielded no results in CAIRN or
OpenEditions. The raw search yielded 142 results. After removing duplicates (n=17), applying exclusion
criteria (Table 1, n=76), removing inaccessible documents (n=2), the final corpus comprises 47 documents
(Figure 1). Documents included in the corpus are marked with an asterisk in the reference list at the end.
Criteria for inclusion and exclusion of texts
As the review was specifically designed to identify publications in the field of AIED dealing with the role of
teachers, the exact phrase “artificial intelligence in education” and the word “teacher*” were identified as
the main inclusion criteria.
revue-mediations.teluq.ca No 16, Autumn 2023
34
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
5
Table 1
Review inclusion and exclusion criteria
Search criteria
Values
Inclusion criteria
Year of publication
1970 … 2022
Artificial intelligence in education
Title, abstract or keyword(s) contain(s): “artificial
intelligence in education” OR “AIED” OR
“intelligence artificielle en éducation”.
The teacher’s role
Title and abstract contain: teacher* OR
enseignant*.
Exclude other definitions of the acronym
AIED (autoimmune inner ear disease)
Title and abstract do not contain: “disease”.
Exclusion criteria (scanning of titles and
abstracts)
Article language
Not in English or French
Type of article
Editorials, data collection policies or calls for
papers
Subject
Deals with medicine or health, deals with a
specific tool without discussing the AIED field
Teachers
Does not address the teacher’s role in AIED
D
November 1
st
2022
142 documents identified in databases
(including 11 cross-referenced
additions)
17 duplicates removed
76 documents excluded for one or more of the following
reasons:
• Discusses a course on AI
• AIED only mentioned as a peripheral element
• Call for papers
• Study of a specific tool without discussing the field
• Does not deal with teachers
125 documents scanned
(title and summary)
47 documents
selected
49 relevant
documents
Figure 1
Corpus document selection process
revue-mediations.teluq.ca No 16, Autumn 2023
35
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
6
Results
The aim of this section is to present the results in relation to the research objective, i.e., to present the way
the roles of AI and teachers are approached in the field of AIED. The first section presents a description of
the corpus, after which the roles of AI and teachers are discussed in turn. For each, results are separated
according to interactions (e.g., role of AI or teacher with teachers, learners, knowledge and class). The
final section presents specific results in relation to the explicit idea of teacher replacement.
Corpus description
A total of 65 coding categories were created and are presented in TTable 2 with the number of
corresponding passages and the number of documents in which this category was used at least once. The
role of AI in education was addressed at least once in 46 documents, and the role of teachers in 29
documents. The majority of documents are theoretical articles published in peer-reviewed journals (n=25).
revue-mediations.teluq.ca No 16, Autumn 2023
36
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
7
Table 2
Coding grid with number of segments and number of documents per code
Code
Seg
Doc
Role of AI
519
46
Working with teachers
106
26
Provide teachers with information on
learners
45
22
Modeling teachers’ work
23
6
Helping with decision-making
19
12
Provide teachers with information on their
practice
15
4
Facilitate collaboration and training among
teachers
4
3
Role with learners
273
43
Provide accurate feedback
56
27
Modeling the learner
55
25
Personalizing learning
40
21
Supporting metacognition
38
18
Assessing learners
20
15
Supporting motivation
16
11
Detecting emotions
15
10
Building a relationship with the learner
12
3
Choosing teaching strategies
9
7
Guiding towards good learning strategies
7
6
Detecting plagiarism
3
2
Identifying at-risk students
2
2
Role in relation to knowledge
105
31
Modeling a knowledge domain
49
19
Choosing content and activities
20
14
Transmitting knowledge
18
14
Producing or enhancing digital educational
resources
15
7
“Drill & practice” exercises
3
3
Role with the class
34
11
Supporting collaborative work
17
8
Classroom and behaviour management
8
1
Fueling discussion between teachers and
learners
7
4
Modeling physical learning spaces
2
2
Code
Seg
Doc
The teacher’s role
132
29
Role with the AI
47
18
Participating in the development of AI
systems
22
10
Interpreting AI data
9
5
Entering data
8
7
Improving or correcting AI representations
5
3
Choosing AI tools
3
3
Role with learners
36
17
Supporting learner motivation
9
6
Representing and getting to know learners
8
7
Making accurate learning diagnoses
6
4
Providing feedback to learners
4
3
Assessing learners
3
3
Guiding towards good learning strategies
3
2
Selecting individual tasks
2
2
Supporting learners’ metacognition
1
1
Role in relation to knowledge
21
11
Instructional planning
11
6
Creating digital educational resources
4
3
Determining elaborate teaching strategies
3
3
Transmitting knowledge
3
3
Role with the class
28
14
Interpreting a unique educational situation
8
3
Relating to the group
7
4
Managing exchanges and collaborative
work
6
5
Creating and maintaining a healthy
classroom climate
3
3
Performing non-goal-oriented peripheral
actions
3
2
Negotiating with students
1
1
Teacher replacement
143
38
Transformation of the teacher’s role
47
24
Benefits of AI for teachers
40
18
Teacher-AI-learner triangle
24
9
Advantages of the teacher over AI
22
13
Differences between an intelligent tutor and a
teacher
10
7
Note: The number of segments coded at the 1
st
and 2
nd
levels is a subtotal. The passages have all been coded at the 3
rd
level. The number of documents at the 3
rd
level is the total number of documents with at least one coded passage. At the 1
st
and 2
nd
levels, this is the number of documents with at least one passage coded in one or more of the 3
rd
level codes.
revue-mediations.teluq.ca No 16, Autumn 2023
37
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
8
Figures 2 to 4 show the distribution of documents by type, country of first author and year. The results are
presented in order of the number of passages coded: role of AI (n=539) and role of the teachers (n=130).
Figure 2
Number of documents by country of first author
Figure 3
Number of documents by type of document
12
9
8
6
4
2 2 2 2
0
2
4
6
8
10
12
14
United
States
United
Kingdom
Other Australia Canada Sweden China Saudi
Arabia
India
25
6
5
4
3
2 2
0
5
10
15
20
25
30
Theoretical
articles
Proceedings Review of
literature
Book or
book's
chapter
Profesional
articles
Empirical
articles
Other
revue-mediations.teluq.ca No 16, Autumn 2023
38
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
9
Figure 4
Number of documents by year of publication
The role of AI
AI’S ROLE WITH RESPECT TO TEACHERS
The main results relating to the role of AI concern decision support and the modeling of teachers’ work.
First, AI can be used to help teachers make decisions. A dozen or so papers discuss decision support, but
most do not go into more detail about the tasks that can be supported. These may include help with
instructional design or the selection of educational resources (Celik et al., 2022), help with diagnosing
learning difficulties (Colbourn, 1985) or help with the whole range of everyday tasks. As such, Timms
(2016) uses the concept of “educational cobot” (p. 703) to describe an intelligent assistant that would
augment the teacher’s capabilities. To support decision-making, AI systems may increase the information
available to teachers. Colbourn (1985) proposes that AI systems can help teachers detect learning
difficulties. More recently, several authors address the idea that AI makes it possible to collate data that
would otherwise be impossible to obtain (Big data paradigm, Cox and Brna, 2016). In the field of learning
analytics, dashboards for teachers make it possible to visualize cumulative information on the learning
traces of learners individually or in groups (Lajoie, 2021). Success prediction tools are also used to provide
teachers with additional information to support their decision-making (Dillenbourg, 2016; Yuskovych-
Zhukovska et al., 2022). According to Humble and Mozelius (2019), an intelligent tutorial system “needs
not to be intelligent but rather designed intelligently to amplify the human tutor’s already existing
intelligence” (p. 5, drawing on Baker, 2016).
Second, AI is also being used to model the work of teachers, with the aim of reproducing their actions more
faithfully. In this respect, recent articles show an intention to go beyond the simple modeling of a teacher’s
role as tutor, and also consider the relational aspect with learners. For example, Celik et al. (2022) report
on studies in which attempts were made to analyze teachers’ verbal communication or movements from
video recordings to model their classroom behaviour. These uses can help to provide a better
understanding of teachers’ gestures so that they can be reproduced by an intelligent tutor, but for the
moment they are mainly used to help future teachers develop a reflexive hindsight on their practice
0
1
2
3
4
5
6
7
8
9
10
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
revue-mediations.teluq.ca No 16, Autumn 2023
39
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
10
(Porayska-Pomsta, 2016). In line with this idea of taking a reflective look at practice, du Boulay (2021)
proposes that systems could be used to help teachers see the gaps between their classroom planning and
the piloting experienced in the classroom.
AI’S ROLE WITH RESPECT TO LEARNERS
The main roles of AI with regard to learners that will be presented are, in order, modeling the learner,
personalizing learning, providing feedback and building a relationship with the learner.
First, AIED systems comprise a learner model, a knowledge domain model and a teaching model (Crovello,
1985; Dede et al., 1985). The learner model is used to store a representation of the learner’s knowledge
(Halff, 1986) and to update it as he or she learns with the system. The learner’s model is then juxtaposed
with the knowledge model or an expert’s model, and is used to infer the learner’s missing or erroneous
knowledge (Brown, 1977; Burton and Brown, 1982). If the learner has integrated the knowledge correctly,
he or she should produce the same responses as the knowledge model (Carbonell, 1970). Since the
beginning of the AIED field in the 1980s, the possibilities for modeling the learner have increased tenfold,
notably through the use of web-based tools that connect data from multiple learners to create more
complex models (Bull and Kay, 2016). What’s more, the proliferation of sensors for collecting data is
leading to greater complexity in learner modeling by incorporating a wider variety (Dillenbourg, 2016).
Learner models now tend to incorporate more and more data on students’ emotional engagement and not
just on their state of knowledge (du Boulay, 2021), as well as longitudinal data (Pinkwart, 2016). These
models underpin any intervention that aims to personalize learning (Kay et al., 2022). According to
Lameras and Arnab (2021), the learner model should include six components: knowledge about the subject
being learned, motivations to learn and expectations of the learning situation, prior experience of different
learning modalities, preferences, social skills and confidence in using an adaptive learning system.
Second, based on the learner model, AIED systems aim to personalize the learning experience. This
personalization can take place in a number of ways, for example by removing or adding options to software
navigation (Brusilovsky and Peylo, 2003). It can also be achieved by adapting content to a person’s
interests (du Boulay, 2021; Khandelwal, 2021) or by gradually adjusting the level of difficulty. According to
du Boulay and Luckin (2016), adaptation can be macro when it targets a group of people, or micro when it
targets individuals.
Third, AIED systems aim to give rapid feedback to learners as mentioned by many authors (e.g., Dede et
al., 1985; Humble and Mozelius, 2019; Kann, 1983; Khan et al., 2022; Stubbs and Piddock, 1985),
sometimes in real time at the moment of performing a procedure. They can also consist of personalized
cues, based on mistakes made by learners, to help them adjust their actions (Brown, 1977). Such feedback
aims to regulate knowledge, support motivation to learn (e.g., Kim and Baylor, 2016; Walker and Ogan,
2016), encourage metacognitive reflection by interrupting the learner to suggest reflections (Dede et al.,
1985). Some systems aim instead for the learner to drive interactions with the system themselves by asking
questions (Jonassen, 2011). Burton and Brown (1982) distinguish the tutor from the “coach” (p. 79), saying
that the tutor acts more formally to supervise specific learning, while the coach is more focused on
encouraging learning in an informal environment where the student has more initiative (e.g., a video game).
Finally, more recently, systems are placing value on building a relationship with the learner. Walker and
Ogan (2016) propose to model these relationships between intelligent tutor and learner:
We propose that AIED systems include designed relationships, or particular care be taken to
construct the socio-motivational relationship between the AIED system and the student. As we
revue-mediations.teluq.ca No 16, Autumn 2023
40
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
11
note above, a growing body of literature suggests that socially-designed interactions with
educational technologies can produce similar outcomes as social interactions amongst teachers
and students or peer collaborators (p. 717).
More and more systems are aiming to detect emotions while performing a task. Various experiments are
taking place, notably to assess the overall attention level of the class based on real-time video analysis
(Raca et al., 2014, cited in Dillenbourg, 2016), for facial recognition or eye tracking (Timms, 2016), or the
classification of emotions (Lameras and Arnab, 2021).
In summary, the role of AI with respect to learners in the field of AIED has mainly been that of intelligent
tutor for personalized learning and rapid feedback when solving a problem. The field is devoting more and
more attention to relational and emotional aspects, and tends to extend the modeling of learners beyond
cognitive aspects, which is made possible in particular by the multiplication of data sources.
THE ROLE OF AI IN RELATION TO KNOWLEDGE
The main roles of AI in relation to knowledge are knowledge modeling and knowledge transmission, which
generally involves the production of learning resources.
First, knowledge modeling is mentioned in 19 of the 47 documents in the corpus, slightly less than learner
modeling (25 documents). Knowledge modeling can be done by creating semantic networks (Halff, 1986)
linking concepts together, or by extrapolating or inferring relationships from a knowledge base (e.g.,
Carbonell, 1970). It is on the basis of this modeling that AIED systems can make instructional decisions
without having been explicitly programmed (Wenger, 1986).
Second, this knowledge modeling done or supported by the AI then enables the transmission of knowledge
to learners. More specifically, the AI’s role may be to select the knowledge to be transmitted (Brusilovsky
and Peylo, 2003) or to demonstrate how to apply a technique (Stubbs and Piddock, 1985). Several papers
mention knowledge transmission indirectly, for example, Ye et al., (2021) referring to Skinner’s learning
machine or systems capable of answering content questions posed by learners (Jonassen, 2011; Stubbs
and Piddock, 1985). Three articles, all from the 1980s, refer to a “Drill & practice” approach (Crovello, 1985;
Kann, 1983; Stubbs and Piddock, 1985). AI systems can also be used to produce or enrich learning
resources. This may involve producing material to meet a student’s particular characteristics (du Boulay,
2021; Porayska-Pomsta, 2016), translating material or generating subtitles automatically (Khandelwal,
2021), or summarizing content (Malik et al., 2019).
THE ROLE OF AI IN THE CLASSROOM
A number of recent documents assign roles to AI in classroom management. There are two such roles:
supporting collaborative work or modeling learning spaces.
First, AI systems can be used to support collaboration among learners by structuring discussions to
maximize their potential (Lameras and Arnab, 2021) or by calculating indicators of engagement in a
collaborative project. For example, Lajoie (2021) reports the use of an online discussion system in which
learners’ contributions are analyzed and related to those of other learners. Dillenbourg (2016, building on
Bachour et al., 2010) reports the use of a table with indicator lights serving as indicators of speaking time
used by individuals. AI can also power interactions between teacher and learners. Open learner models
also fulfil this function, as they provide information about learning and learners (Kay et al., 2022).
revue-mediations.teluq.ca No 16, Autumn 2023
41
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
12
Second, in connection with the classroom, AI could increasingly be used to model physical learning spaces:
[…] while AIED initially aimed at modelling the contents and the learner, a challenge for the future
of AIED is to model educational spaces, i.e., the physical space and the diverse actors who inhabit
this space, in order to make education more effective. We conceptualized this evolution by defining
a third circle of usability (Dillenbourg et al., 2011), where the user is not an individual (first circle)
or a team (second circle) but the entire classroom is viewed as a physical and sociological system.
(Dillenbourg, 2016, p. 548)
Modeling the classroom means going beyond personalizing learning on a purely individual basis and
integrating, as Dillenbourg points out, knowledge about the group and how it functions.
To sum up, the role of AI in the classroom is little discussed in the corpus. Nevertheless, it seems that the
role of AI in the classroom is developing more in the AIED field today than in its early days (26 of the 34
passages coded in this category come from documents published after 2010), which is consistent with the
growing interest in relational and affective aspects presented in the section on the role of AI in learners.
This role may involve structuring exchanges or modeling both physical and digital learning spaces
(Dillenbourg, 2016).
The teacher’s role
THE TEACHER’S ROLE WITH RESPECT TO AI
The role of the teacher regarding AI is rarely mentioned in the corpus (47 passages in 18 documents).
When it is, it is essentially in two respects: either as a passive user of the systems, or as an active user
invited to configure a system or enter data.
As passive users of AIED systems, teachers may be required first and foremost to interpret information
provided by AI: “teachers need to understand the results of intelligent analysis based on teaching situations
and pedagogical theories, compare the gap between students’ achievements, identify students learning
needs, predict risks at academic failure, and even discovering new rules for AIED” (Liu and Li, 2022, p. 39).
This role can be fulfilled to a greater or lesser extent depending on the teacher’s level of data literacy
(Howard et al., 2022). As such, even passive use may involve teachers having to develop knowledge to
integrate the use of AI into their teaching practice (Liu and Li, 2022; Pinkwart, 2016).
As active users, teachers may have to contribute to producing this data. Bull and Kay (2016) assert that
the learner’s model can either be controlled entirely by the system, or controlled jointly by the teacher and
the system. Thus, in addition to interpreting data, the teacher could be required to enter or modify data to
complexify or correct the learner model (Bull and Kay, 2016), including information relating to behaviours
that would escape the digital traces (Celik et al., 2022). According to Liu and Li (2022), teachers also have
an ethical responsibility towards AI: “Teachers need to have the correct value judgment in deploying
intelligent technology to promote students’ learning and well-being, understand the potential risks of AIED,
and handle the ethical issues in a prudent and responsible manner” (p. 37).
revue-mediations.teluq.ca No 16, Autumn 2023
42
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
13
THE TEACHER’S ROLE WITH RESPECT TO LEARNERS
In the field of AIED, there is little discussion about the role of the teacher with respect to learners, and
when it is, it often illustrates the gap between what AIED systems do and how much remains to be done
to imitate or even replace the teacher. The teacher’s main roles regading learners are to support
motivation, accurately diagnose obstacles to comprehension and get to know the learners.
First, teachers are presented as essential players in supporting learner motivation. Liu and Li (2022)
mention that the emotional work of teachers is essential to create a positive atmosphere that encourages
the pleasure of learning and self-improvement. To create this atmosphere, Timms (2016) evokes the
importance of teachers demonstrating a genuine and personal interest in learners beyond the subject
matter.
Second, several authors have also mentioned the teacher's role in identifying learning pitfalls in complex
situations or with fragmentary, disorganized and partial information (Les et al., 1999), an idea also
mentioned by (Carbonell, 1970): “Human teachers sometimes try to understand the nature of their
students’ confusions and problems, but at least as often, they go into explanatory and remedial sequences
without a full understanding of the reasons for the students’ errors.” (pp. 198-199). Du Boulay (2021)
stresses the importance of teachers in supporting metacognition to consolidate learning.
Finally, to fulfill these motivational support and pitfall identification roles, teachers need to know their
learners and maintain mental representations of them (Goodyear et al., 1989). This aspect has been
referred to by several authors since the early days of the field, e.g., Crovello (1985) mentioning that
teachers must have knowledge about each individual learner. Cumming et al. (1997) refer to cognitive,
affective and social knowledge, some of which is dynamic and changes according to the situation, while
others are long-term (e.g., learners’ personality traits). Kay et al. (2022) speak of the “model of the learner
in the mind of the teacher” and the “model of a set of learners” (p. 5).
In short, the teacher’s role with respect to learners is essentially relational, and it is from this base that
tasks related to learning support are carried out.
THE TEACHER’S ROLE IN RELATION TO KNOWLEDGE
The most discussed role of the teacher in relation to knowledge is instructional planning. Ahmad et al.
(2022) speak of curriculum development or the creation of lesson plans, and Liu and Li (2022) even
consider that “human teachers play irreplaceable roles in curriculum and creative professional practice
compared with AI teachers” (p. 35). Teachers are required to create video resources and prepare lectures
(Khandelwal, 2021) and develop resources using AIED tools (Yuskovych-Zhukovska et al., 2022).
Teachers are also responsible for transmitting knowledge and steering elaborate learning activities in real
time, an aspect that until recently was neglected in the field of AIED according to Dillenbourg (2016): “The
role of teacher during runtime did not receive much attention for two decades, but this changed a lot over
the last decade, with the growing interest for the orchestration of computer-enhanced learning activities”
(p. 555).
revue-mediations.teluq.ca No 16, Autumn 2023
43
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
14
THE TEACHER’S ROLE IN RELATION TO THE CLASS
A few passages about learners described the teacher’s work in relation to a group rather than individuals.
The teacher enters into a relationship with a group by calling on communication skills that are not
necessarily specific to teaching:
While there are some specialized tactics that human teachers apply effectively, good teaching
derives from the conversational and social interactive skills used in everyday settings such as
listening, eliciting, intriguing, motivating, cajoling, explaining, arguing, persuading, enthralling,
leading, pleading and so on. (du Boulay and Luckin, 2016, p. 396.)
In the same vein, Porayska-Pomsta (2016) refers to the need for “agile adaptation” (p. 685), also described
as “teacher immediacy” by Walker and Ogan (2016, pp. 716-717), which includes spontaneous gestures
such as smiling, eye contact, gesticulation or the use of common references.
Teacher replacement
Given that this literature review finds its relevance in fears, founded or unfounded, that AI could replace
the teacher, we thought it useful to code the passages in the corpus that discussed precisely this idea.
References to the idea of replacement are generally very brief, for example Robertson, (1976) who states
that “the idea is that such systems may make conventional teaching methods more effective, not that they
should replace them” (p. 437). Conversely, Brusilovsky and Peylo (2003), referring to the field of computer-
assisted instruction, assert that these systems “were intended to replace all or part of traditional classroom
instruction” (p. 163). Colbourn (1985, p. 521) states that in some cases the system acts like a teacher, but
in most cases it acts like a tutor to accompany learners in discovering information or laws for themselves.
Dede et al. (1985) are more straightforward, speaking of the “potential for direct substitution of teacher
activities” (p. 89) and wondering whether the future of the field will involve an intention to automate or
accompany teacher activities. Kann (1983) argues that programs developed in AIED attempt to replicate
the characteristics of the best teachers, such as engaging in two-way communication with learners and
taking account of their interest in whether to pursue certain learning. More recently, Edwards and Cheok
(2018) speak of AIED as a solution to the shortage of manpower in the field. Despite such formulations,
which display the intention of replacing teachers in some of their activities, Dillenbourg (2016) asserts:
“despite a few discordant voices (‘neo-Illich’ gurus), educational technology researchers have never
believed that their technology would suppress the need for teachers in formal education” (p. 555). In his
opinion, however, few studies have been carried out on the roles of teachers in systems developed in the
field of AIED. Humble and Mozelius (2019) ask the question directly: is the aim of the AIED to support
teachers or to replace them?
A number of authors point to the transformation of the role of teachers, as they become more involved in
steering high-level activities (du Boulay, 2021). Time spent with students and their role may also change
(du Boulay, 2021). They may also spend time participating in the co-design of AIED systems (Porayska-
Pomsta, 2016). According to Yuskovych-Zhukovska et al. (2022):
[…] AI is consistently and confidently changing the role of teachers. AI can perform tasks such as
assessment, can help learners improve learning, and can even replace real learning. AI systems
can be a source of expertise to which students can direct their questions, or even take the
teacher’s place for the basic materials of the course. However, in most cases, AI will only change
the role of the teacher to the role of facilitator (p. 350).
revue-mediations.teluq.ca No 16, Autumn 2023
44
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
15
In short, the idea of teacher replacement never seems to have been explored in depth in the field of AIED.
When it is discussed, it’s mostly in a peripheral way to assert that it’s not possible, with contradictions in
terms of the goal pursued (to replace the teacher or not). A few authors do, however, elaborate on the idea
of transforming the role of the teacher in the context of the accelerated development of AIED.
Discussion
Analysis of the results shows that the roles assigned to AI in the field of AIED are those that normally fall
to the teacher. So, even if the expression teacher replacement elicits a number of fears and even if this
objective is, for the time being, unattainable, the fact remains that it seems to be one of the scientific
ambitions of the field—an ambition all the more difficult to detect as the opposite is sometimes stated. This
can be seen in the comments of several authors, who are seeking to model the teacher’s role as well as
possible, including his or her emotional work and the management of social interactions, to better design
AIED systems. Given this observation and the lack of development in the thinking about the transformation
of the teacher’s role in the context of increasingly complex and widespread AIED, it seems essential to us
that educational systems clarify the desired interactions between the different players—the AI, the teacher
and the learner. Failing this, there is a risk that new actions needed to regulate learning will be taken neither
by the AI nor by the teacher, and that informal actions that have until now been taken by the teacher will
be abandoned through gradual delegation to AIED systems.
Some authors have proposed using a teacher-learner-AI triangle to conceptualize roles and their
interactions, starting with Cumming et al. (1997). According to du Boulay (2021), the study of interactions
among those three components has led to a real appreciation of the importance of the teacher in the
educational environment. Humble and Mozelius (2019), drawing on several sources, refer to a values
problem that can potentially hinder the successful deployment of AIED, for example, when the values that
support the development of strong AI are misaligned with those of the people who are to use it, and in this
regard evoke the importance of “human-compatible AI” (p. 2). Celik et al. (2022) proposed a loop of
interactions between teachers and AI in which teachers set assessment criteria, review AI decisions,
document technical issues, feed learner data into the systems, and the AI carries out assessments, tracks
student progress and informs teachers’ planning. As the review brought to light the relationships among
AI, teachers, learners and knowledge, we also propose to conceptualize those interactions on the basis of
Houssaye’s didactic triangle (1988), originally published in 1988 and widely mobilized in the field of
education.
The didactic triangle has already been revisited several times to incorporate ICT or the computer, but never
to our knowledge to incorporate AI specifically. Faerber (2003) was the first, to our knowledge, to propose
an update of the didactic triangle by integrating a technological pole. Essentially, he starts from the
observation that the relationships identified by Houssaye (1988) are modified when teaching-learning takes
place via a virtual environment: “l’environnement virtuel d’apprentissage est un intermédiaire à la fois
fonctionnel, matériel, logiciel entre les pôles” (p. 202). According to Yassine (2010), the role of the
computer in the didactic triangle depends on how it is used. It can act as a “ordinateur enseigné”, in which
case the student programs the computer, as was already the case in Papert’s studies with the LOGO
learning environment. It can also be a “tuteur”, i.e., it leads learners “through a number of trial-and-error
steps to help them recognize their shortcomings and acquire knowledge without teacher intervention” (par.
4.1.2). And it can also be an “outil” i.e., instrumentalized by the teacher as a support for certain types of
learning, for example, to present information. According to these different ways of looking at the computer,
its role is more or less active in the didactic triangle. We propose that AI, as it is conceptualized today, and
revue-mediations.teluq.ca No 16, Autumn 2023
45
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
16
in continuity with research in the field of AIED, aims to play an increasingly active role within the didactic
triangle. Indeed, we have observed a gradual shift between the early writings in the field and those of
today, with the former talking more about AI systems or software using AI, while the latter speak more
generally of artificial intelligence almost giving the impression of personification. AI is less and less seen
as passive and instrumental, given the complexity of the decisions it can make. This, incidentally, echoes
a distinction made by Wenger (1986) between computer-aided instruction (CAI) and AIED: systems
developed in the field of AIED can make instructional decisions without having been specifically
programmed to do so, unlike those developed in the field of CAI.
The didactic triangle has also been revisited by Lombard (2007) to develop the ICT tetrahedron. According
to Lombard, “le maître en classe est très souvent ignoré ou son rôle minimisé ” (par. 30). In his view, many
educational technologies, such as educational games, are often designed without regard for the role of the
teachers, and rather with a view to doing something that normally falls to them. In this case, the uses are
of the order of “l’alternance [entre le professeur] et le dispositif cyber-prof” (par. 37) and do not fall within
the scope of pedagogical integration. Even if we are not talking specifically about AIED tools, the same
question can be asked of the latter: should there simply be alternation between the teaching provided by
the teacher and that provided by AI, or should there be pedagogical integration of the AI by the teacher?
A lack of attention to the desired interactions between a technological device and the teacher can lead to
“des conflits sournois” (par. 45), for example, a reduction in the quality of the pedagogical relationship, or
even its abandonment by the introduction of an intermediary. On the subject of teacher-device
collaboration, Lombard states that “les – plutôt rares – usages des technologies où une collaboration
efficace s’établit entre les 2 pôles pédagogiques que nous avons pu observer, semblent majoritairement
des usages où le [dispositif] joue un rôle très peu intrusif sur le plan de la relation pédagogique” (p. 23).
We therefore propose to re-use Faerber’s tetrahedron of ICT in education (2003) and re-discussed by
Lombard (2007) to examine the new reality arising from the growing permutation of AI in the educational
context. Unlike Lombard (2007), we argue that the tetrahedron should not, or should no longer, become
entangled in studying uses of computers in the same way, i.e., where they are instrumentalized by the
teacher or learner, and those where AI actively transforms the learning situation. At the very least, such
uses would be best studied on the edges of the tetrahedron for the mediation they operate on or among
actors, but should not occupy a vertex in their own right. In fact, Faerber (2003) explicitly states that he
“[n’a pas conféré à l’environnement virtuel] un statut de pôle au même titre que l’apprenant ou le savoir”
(p. 202). Explicitly integrating AI at the apex of the tetrahedron devoted to what Lombard calls the
“Dispositif Cyber-prof” raises new research questions that have not traditionally been part of the AIED field.
These questions, which touch directly on the idea of interaction between teachers and AI, and the
confronting idea of replacing teachers with a machine, need to be asked and studied if satisfactory answers
are to guide the efforts of teachers and educational systems more generally. The teacher should not
continue to do what AI does better, and AI should not be used to do what we don’t yet understand about
the teacher’s role.
Figure 5 presents questions that we believe should inform the design of AIED systems, as well as research
in the field. Such an approach also involves a move away from seeing AIED systems as tools to aid
teaching and learning, and to see them as full players in the process, a change we believe is essential in
view of advances made in the field and in preparation for those to come.
revue-mediations.teluq.ca No 16, Autumn 2023
46
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
17
Figure 5
Proposed framework for thinking about AI-teacher-learner interactions based on the tetrahedron of ICT in
education by Faerber (2003).
Limitations
Despite our efforts to include as many documents as possible by searching several databases and
integrating cross-references, it is possible that some relevant documents were not found, particularly those
not indexed in digital databases. Document coding was carried out by only one of the authors, but the grid
was adjusted by the researchers at working meetings during the analysis process. Finally, as the articles
are mainly theoretical, the roles of AI and the teacher are primarily anticipated rather than observed.
Conclusion
This literature review was based on a corpus of 48 documents evoking the role of the teacher in the field
of AIED. Through an inductive analysis, it brought to light the relationships among teachers, learners, AI
and knowledge as they are conveyed in the field. The main finding is that the roles of the teacher and the
learner are given little attention in the field, compared with that of AI. Despite repeated claims that AI is not
intended to replace the teacher, the actions delegated to it tend to show that the aims are to automate
tasks that normally fall to the teacher (e.g., assessing learners, supporting motivation, providing accurate
feedback), even if this ambition is not achievable in the foreseeable future. Given advances in AI and the
growing complexity of tasks that can be automated, it seems essential to better conceptualize roles to
revue-mediations.teluq.ca No 16, Autumn 2023
47
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
18
ensure that essential teacher tasks that are incompletely modeled are not abandoned to AI (e.g., informal
actions that are nonetheless important). Similarly, given the advances in the field of emotion detection and
even classroom activity monitoring, it seems essential to further study the effects of replacing the teacher
with AI, not only on cognitive aspects, but also on behavioural and affective ones. Based on an adaptation
of Houssaye’s didactic triangle (1988) and Faerber’s ICT tetrahedron (2003), we have proposed questions
that could guide research and design in the field of AIED, taking into account the roles of the learner, the
teacher and AI.
Acknowledgements
Thanks to Didier Paquelin, professor at Université Laval, for his insight into the didactic triangle.
This research was funded by a doctoral scholarship from the Social Sciences and Humanities Research
Council of Canada.
List of references
1
*Ahmad, S. F., Alam, M. M., Rahmat, M. K., Mubarik, M. S., and Hyder, S. I. (2022). Academic and Administrative Role of
Artificial Intelligence in Education. Sustainability, 14(3), 1101. https://doi.org/10.3390/su14031101
Baker, R. S. (2016). Stupid tutoring systems, intelligent humans. International Journal of Artificial Intelligence in Education,
26(2), 600-614.
*Brown, J. S. (1977). Uses of artificial intelligence and advanced computer technology in education. In J. Seidel and L.
Rubin (eds.), Computers and Communications: Implications for Education. Academic Press.
*Brusilovsky, P., and Peylo, C. (2003). Adaptive and Intelligent Web-based Educational Systems. International Journal of
Artificial Intelligence in Education, 13, 156-159.
Bulger, M. (2016). Personalized Learning: The Conversations We’re Not Having. Data & Society, 29.
*Bull, S., and Kay, J. (2016). SMILI☺: A Framework for Interfaces to Learning Data in Open Learner Models, Learning
Analytics and Related Fields. International Journal of Artificial Intelligence in Education, 26(1), 293-331.
https://doi.org/10.1007/s40593-015-0090-8
Burton, R. R., and Brown, J. S. (1982). An investigation of computer coaching for informal learning activities. In D. Sleeman
and J. S. Brown, Intelligent tutoring systems (pp. 79-98). Academic Press.
*Carbonell, J. (1970). AI in CAI: An Artificial-Intelligence Approach to Computer-Assisted Instruction. IEEE Transactions on
Man Machine Systems, 11(4), 190-202. https://doi.org/10.1109/TMMS.1970.299942
*Celik, I., Dindar, M., Muukkonen, H., and Järvelä, S. (2022). The Promises and Challenges of Artificial Intelligence for
Teachers: A Systematic Review of Research. TechTrends, 66(4), 616-630.https://doi.org/10.1007/s11528-022-
00715-y
*Colbourn, M. J. (1985). Applications of artificial intelligence within education. Computers & Mathematics with Applications,
11(5), 517-526. https://doi.org/10.1016/0898-1221(85)90054-9
Corbin, J. M., and Strauss, A. L. (2015). Basics of qualitative research: Techniques and procedures for developing
grounded theory (4
th
ed.). SAGE.
1
Documents included in the corpus are marked with an asterisk.
revue-mediations.teluq.ca No 16, Autumn 2023
48
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
19
* Cox, R., and Brna, P. (2016). Twenty Years on: Reflections on “Supporting the Use of External Representations in
Problem Solving”.... International Journal of Artificial Intelligence in Education, 26(1), 193-204-.
https://doi.org/10.1007/s40593-015-0054-z
*Crovello, T. J. (1985). Computers, education and artificial intelligence. ACM ‘85: Proceedings of the 1985 ACM annual
conference on The range of computing: mid-80’s perspective.
*Cumming, G., Cropp, S., and McDougall, A. (1997). Learner Modelling: Lessons from Expert Human Teachers. In B. du
Boulay and R. Mizoguchi (eds.), Proceedings of the 8
th
World Conference on Artificial Intelligence in
Education―Knowledge and Media in Learning Systems (AI-ED 97) (pp. 577-579). IOS Press.
*Dede, C. J., Zodhiates, P. P., and Thompson, C. L. (1985). Intelligent Computer-Assisted Instruction: A Review and
Assessment of ICAI Research and Its Potential for Education. Educational Technology Center.
*Dillenbourg, P. (2016). The Evolution of Research on Digital Education. International Journal of Artificial Intelligence in
Education, 26(2), 544-560. https://doi.org/10.1007/s40593-016-0106-z
*du Boulay, B. (2021). Jim Greer’s and Mary Mark’s Reviews of Evaluation Methods for Adaptive Systems: A Brief
Comment about New Goals. International Journal of Artificial Intelligence in Education, 31(3), 622-635.
https://doi.org/10.1007/s40593-020-00198-z
*du Boulay, B., and Luckin, R. (2016). Modelling Human Teaching Tactics and Strategies for Tutoring Systems: 14 Years
On. International Journal of Artificial Intelligence in Education, 26(1), 393-404.
https://doi.org/10.1007/s40593-015-0053-0
*Edwards, B. I., and Cheok, A. D. (2018). Why Not Robot Teachers: Artificial Intelligence for Addressing Teacher Shortage.
Applied Artificial Intelligence, 32(4), 345-360. https://doi.org/10.1080/08839514.2018.1464286
Faerber, R. (2003). Groupements, processus pédagogiques et quelques contraintes liés à un environnement virtuel
d’apprentissage. Environnements informatiques pour l’apprentissage humain, 200-210.
*Goodyear, P., Oosthoek, H., and Vroeijenstijn, T. (1989). Expert systems and intelligent tutoring—Some issues in the
engineering of pedagogic knowledge. In Higher Education and New Technologies (pp. 45-51). Pergamon.
*Halff, H. M. (1986). Instructional Applications of Artificial Intelligence. Educational Leadership: Journal of the Department
of Supervision and Curriculum Development, 24-31.
Holmes, W., Porayska-Pomsta, K., Holstein, K., Sutherland, E., Baker, T., Shum, S. B., Santos, O. C., Rodrigo, M. T.,
Cukurova, M., Bittencourt, I. I., and Koedinger, K. R. (2021). Ethics of AI in Education: Towards a Community-Wide
Framework. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-021-00239-1
Houssaye, J. (1988). The educational triangle. Peter Lang.
*Howard, S. K., Swist, T., Gasevic, D., Bartimote, K., Knight, S., Gulson, K., Apps, T., Peloche, J., Hutchinson, N., and
Selwyn, N. (2022). Educational data journeys: Where are we going, what are we taking and making for AI?
Computers and Education: Artificial Intelligence, 3, 100073. https://doi.org/10.1016/j.caeai.2022.100073
*Humble, N., and Mozelius, P. (2019). Artificial Intelligence in Education—A Promise, a Threat, or a Hype? Proceedings of
the European Conference on the Impact of Artificial Intelligence and Robotics, 149-156, Oxford, UK.
*Humble, N., and Mozelius, P. (2019). Teacher-supported AI or AI-supported teachers? Proceedings of the European
Conference on the Impact of Artificial Intelligence and Robotics, Oxford, UK.
*Jonassen, D. H. (2011). Ask Systems: Interrogative access to multiple ways of thinking. Educational Technology
Research and Development, 59(1), 159-175. https://doi.org/10.1007/s11423-010-9179-9
*Kann, L. (1983). Artificial intelligence and its implications for education. Canadian journal of educational communication,
12(3).
*Kay, J., Bartimote, K., Kitto, K., Kummerfeld, B., Liu, D., and Reimann, P. (2022). Enhancing learning by Open Learner
Model (OLM) driven data design. Computers and Education: Artificial Intelligence, 3, 100069.
https://doi.org/10.1016/j.caeai.2022.100069
*Khan, M. A., Khojah, M., and Vivek (2022). Artificial Intelligence and Big Data: The Advent of New Pedagogy in the
Adaptive E-Learning System in the Higher Educational Institutions of Saudi Arabia. Education Research
International, 2022, 1-10. https://doi.org/10.1155/2022/1263555
revue-mediations.teluq.ca No 16, Autumn 2023
49
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
20
*Khandelwal, K. (2021). Application of AI to Education during the Global Crisis. Review of international geographical
education, 7, 3204-3212. https://rigeo.org/menu-script/index.php/rigeo/article/view/2583
*Kim, Y., and Baylor, A. L. (2016). Research-Based Design of Pedagogical Agent Roles: A Review, Progress, and
Recommendations. International Journal of Artificial Intelligence in Education, 26(1), 160-169.
https://doi.org/10.1007/s40593-015-0055-y
Koehler, M. J., and Mishra, P. (2009). What Is Technological Pedagogical Content Knowledge? Contemporary Issues in
Technology and Teacher Education, 9(1), 11.
*Lajoie, S. P. (2021). Student Modeling for Individuals and Groups: The BioWorld and HOWARD Platforms. International
Journal of Artificial Intelligence in Education, 31(3), 460-475. https://doi.org/10.1007/s40593-020-00219-x
*Lameras, P., and Arnab, S. (2021). Power to the Teachers: An Exploratory Review on Artificial Intelligence in Education.
Information, 13(1), 14. https://doi.org/10.3390/info13010014
*Lelouche, R. (2000). The various benefits of education from computers: An integrative review. Proceedings International
Workshop on Advanced Learning Technologies. IWALT 2000. Advanced Learning Technology: Design and
Development Issues, 7-7. https://doi.org/10.1109/IWALT.2000.890550
*Les, J., Cumming, G., and Finch, S. (1999). Agent systems for diversity in human learning. In S. P. Lajoie and M. Vivet
(eds.), Artificial intelligence in education—Open Learning Environments: New Computational Technologies to
Support Learning, Exploration and Collaboration (pp. 13-20).
*Liu, X., and Li, Y. (2022). Redefining Teacher Qualification in the Artificial Intelligence Era: A Professional Capital
Perspective. Proceedings of the 5
th
International Conference on Big Data and Education, 35-39.
https://doi.org/10.1145/3524383.3524405
Lombard, F. (2007). Du triangle de Houssaye au tétraèdre des TIC : comprendre les interactions entre les savoirs
d’expérience et ceux de recherche. In B. Charlier and D. Peraya (Eds.), Transformation des regards sur la
recherche en technologie de l’éducation (pp. 137-154). De Boeck Supérieur; Cairn.info.
https://doi.org/10.3917/dbu.charl.2007.01.0137
*Malik, G., Tayal, D. K., and Vij, S. (2019). An Analysis of the Role of Artificial Intelligence in Education and Teaching. In P.
K. Sa, S. Bakshi, I. K. Hatzilygeroudis, and M. N. Sahoo (eds.), Recent Findings in Intelligent Computing
Techniques (Vol. 707, pp. 407-417). Springer Singapore. https://doi.org/10.1007/978-981-10-8639-7_42
Marrhich, A., Lafram, I., Berbiche, N., and El Alami, J. (2021). Teachers’ Roles in Online Environments: How AI Based
Techniques Can Ease the Shift Challenges from Face-to-Face to Distance Learning. International Journal of
Emerging Technologies in Learning (IJET), 16(24), 244-254. https://doi.org/10.3991/ijet.v16i24.26367
*McLaren, B. M., and Scheuer, O. (2010). Supporting Collaborative Learning and E-Discussions Using Artificial Intelligence
Techniques. International Journal of Artificial Intelligence in Education, 20(1).
https://doi.org/10.3233%2FJAI-2010-0001
Minsky, M. (1991). Logical Versus Analogical or Symbolic Versus Connectionist or Neat Versus Scruffy. AI Magazine,
12(2), 34-51.
Mubin, O., Stevens, C. J., Shahid, S., Mahmud, A. A., and Dong, J.J. (2013). A review of the applicability of robots in
education. Technology for Education and Learning, 1(2013), 1-7.
https://doi.org/10.2316/Journal.209.2013.1.209-0015
*Naughton, J. (1987). Artificial intelligence and education and training. In Artificial Intelligence (pp. 77-82). Elsevier.
https://doi.org/10.1016/B978-0-08-034112-5.50012-7
*Pervez, S., and Alandjani, G. (2018). 21
st
Century Educational Requirements and Teaching Strategies for Competing with
the Cyborgs. International E-Journal of Advances in Social Sciences, 4(11), 10.
*Pinkwart, N. (2016). Another 25 Years of AIED? Challenges and Opportunities for Intelligent Educational Technologies of
the Future. International Journal of Artificial Intelligence in Education, 26(2), 771-783.
https://doi.org/10.1007/s40593-016-0099-7
*Porayska-Pomsta, K. (2016). AI as a Methodology for Supporting Educational Praxis and Teacher Metacognition.
International Journal of Artificial Intelligence in Education, 26(2), 679-700.
https://doi.org/10.1007/s40593-016-0101-4
revue-mediations.teluq.ca No 16, Autumn 2023
50
Revue internationale sur le numérique en éducation et communication
© Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
Authors. This work is distributed under a Creative Commons Attribution 4.0 International license.
revue-mediations.teluq.ca | N
o
16, 2023
21
Renz, A., and Vladova, G. (2021). Reinvigorating the Discourse on Human-Centered Artificial Intelligence in Educational
Technologies. Technology Innovation Management Review, 11(5), 5-16. https://doi.org/10.22215/timreview/1438
*Richer, M. H. (1985). Applications of artificial intelligence in education—A personal view. The Physiologist, 28(5), 428-431.
Rhoades, E. (2011). Literature reviews. The Volta Review, 111(3), 353-368.
*Robertson, M. (1976). Artificial intelligence in education. Nature, 262, 435-437.
Romiszowski, A. J. (1987). Artificial Intelligence and Expert Systems in Education: Potential Promise or Threat to
Teachers? Educational Media International, 24(2), 96-104. https://doi.org/10.1080/0952398870240208
*Salem, A.-B. M. (2000). The Potential Role of Artificial Intelligence Technology in Education. Proceedings of the
International Conference on Technology in Mathematics Education, 178-185.
Saltman, K. J. (2020). Artificial intelligence and the technological turn of public education privatization: In defence of
democratic education. London Review of Education, 18(2). https://files.eric.ed.gov/fulltext/EJ1297432.pdf
Self, J. (2016). The Birth of IJAIED. International Journal of Artificial Intelligence in Education, 26(1), 4-12.
https://doi.org/10.1007/s40593-015-0040-5
*Sleeman, D., and Brown, J. S. (1982). Intelligent tutoring systems. Academic Press.
*Stubbs, M., and Piddock, P. (1985). Artificial Intelligence in Teaching and Learning: An Introduction. PLET: Programmed
Learning & Educational Technology, 22(2), 150-157. https://doi.org/10.1080/1355800850220207
*Timms, M. J. (2016). Letting Artificial Intelligence in Education Out of the Box: Educational Cobots and Smart Classrooms.
International Journal of Artificial Intelligence in Education, 26(2), 701-712.
https://doi.org/10.1007/s40593-016-0095-y
*Walker, E., and Ogan, A. (2016). We’re in this Together: Intentional Design of Social Relationships with AIED Systems.
International Journal of Artificial Intelligence in Education, 26(2), 713-729.
https://doi.org/10.1007/s40593-016-0100-5
*Wenger, E. (1986). Artificial intelligence and tutoring systems: Computational approaches to the communication of
knowledge. Morgan Kaufmann.
Yassine, J. (2010). The didactic triangle seen in the light of the introduction of ICT. EpiNet, 128.
https://edutice.archives-ouvertes.fr/edutice-00560705/file/a1011e.htm
*Ye, R., Sun, F., and Li, J. (2021). Artificial Intelligence in Education: Origin, Development and Rise. In Liu, X.-J., Nie, Z.,
Yu, J., Xie, F., Song, R. (eds.), Intelligent Robotics and Applications (Vol. 13016, pp. 545-553). Springer
International Publishing. https://doi.org/10.1007/978-3-030-89092-6_49
*Yuskovych-Zhukovska, V., Poplavska, T., Diachenko, O., Mishenina, T., Topolnyk, Y., and Gurevych, R. (2022).
Application of Artificial Intelligence in Education. Problems and Opportunities for Sustainable Development. BRAIN.
Broad Research in Artificial Intelligence and Neuroscience, 13(1Sup1), 339-356.
https://doi.org/10.18662/brain/13.1Sup1/322