Virtual Assistants for Learning: A Systematic Literature Review
Regina Gubareva
1
and Rui Pedro Lopes
2 a
1
Taraz State University, Kazakhstan
2
Research Center for Digitalization and Industrial Robotics, Instituto Polit
´
ecnico de Braganc¸a, Portugal
Keywords:
Virtual Assistant, Learning Management System, Machine Learning, Higher Education, Systematic Literature
Review.
Abstract:
A problem of students’ motivation, engagement and declining interest in the learning process has always ex-
isted, contributing to increasing failures and dropouts. This is particularly important among first year students.
Freshmen often have difficulties with time management, how to prioritize tasks, and how to study at the uni-
versity. Because of the increasing number of higher education students, it is impossible to provide individual
tutoring and support to every student, to help them manage this first-year indefiniteness and later difficul-
ties. Recent developments in the area of information technology, software engineering, artificial intelligence,
machine learning, and big data creates the opportunity for personalized, flexible and adaptable learning envi-
ronment, accessible anytime, anywhere. One such example is a virtual assistant, a tool that provides assistance
to usually boring or repetitive daily activities. In education, a virtual assistant can help organising the study
process, manage time, increasing motivation and engagement in the study process. This paper performs a sys-
tematic literature review of the use of virtual assistants in higher education. It focuses on the technology that
powers them, their features and their impact in the learning process, motivation and productivity, according to
the authors.
1 INTRODUCTION
Education is one of the most important aspects of hu-
man development, greatly influencing the path of pro-
fessional development and success (Mesquita et al.,
2015; Mesquita et al., 2014). The increasing need
for training at the higher-education level, that con-
tributes to the scientific and technological qualifica-
tion of youth and adults, towards the “creative, inno-
vative and competitive development, with high pro-
ductivity standards” (Correia and Mesquita, 2006, p.
166) is a huge challenge that higher education institu-
tions face. The admission of students with heteroge-
neous profiles, with different motivations, academic
and social background, with diverse influences cre-
ate important asymmetries, requiring adequate analy-
sis of their characteristics and profiles to optimize the
academic success.
Higher education is, thus, redefining its paradigm,
moving from a traditional lectures perspective to an
active and emancipating learning process. This re-
quires not only the understanding of the audiences,
but also permanent reflection and cooperation be-
a
https://orcid.org/0000-0002-9170-5078
tween all the involved actors. In fact, although the
integration of students in higher education foster their
intellectual and ethical behaviour, not many manage
to achieve higher levels in this development. This
requires considerable effort from institutions to help
students to achieve higher intellectual, ethical and
professional development.
Adequate pedagogical and training strategies,
such as the involvement of teachers in their pedagogi-
cal training (Rosado-Pinto, 2008), tutoring, project-
based learning (Veiga Sim
˜
ao et al., 2008; Sim
˜
ao
et al., 2002) have contributions to the integration and
learning of students in higher education (Lopes and
Mesquita-Pires, 2013).
To a higher degree, the quality of the learning
process depends on student motivation and involve-
ment (Hanus and Fox, 2015), even when facing mas-
sification of education, declining completion rates,
and, sometimes, incoherence of courses. This, further
associated with the students’ difficulty in managing
study time, contributes to increasing dropout and low
academic success. The workload required for a stu-
dent to achieve the objectives of a particular course
is dependent on their work skills, the level of the ob-
jectives and contents of each course and the teaching-
Gubareva, R. and Lopes, R.
Virtual Assistants for Learning: A Systematic Literature Review.
DOI: 10.5220/0009417600970103
In Proceedings of the 12th International Conference on Computer Supported Education (CSEDU 2020) - Volume 1, pages 97-103
ISBN: 978-989-758-417-6
Copyright
c
2020 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
97
learning methods. The process puts great emphasis on
the student, encouraging higher education institutions
and academic staff to place students at the centre of
their thinking and to help them manage their expec-
tations and be able to consciously and constructively
design their learning paths throughout their higher ed-
ucation experience (Lopes et al., 2019; Tenorio et al.,
2018).
In this framework, the students’ autonomous work
assumes a fundamental value in learning. It is impor-
tant that they know and recognize the objectives, what
is being studied, how to define the work tasks and pri-
orities, how to use and to enjoy the numerous infor-
mation resources, how to write summaries and to pre-
pare reading sheets and reports. In other words, they
need to be persistent and manage to take the work to
the end. This suggests that students can benefit from
aid to help them manage all the complexity associ-
ated with the what (what they need to do), when (the
amount of time, the deadlines, the schedule organiza-
tion) and how (the content, exercises and procedures)
to learn. This is where a virtual assistant can be use-
ful.
This paper explores, through a systematic litera-
ture review, the use and importance of virtual assistant
in the motivation and autonomy of higher education
students.
2 METHODOLOGY
The main objective of this literature review is to try to
understand the role of virtual assistants in higher ed-
ucation, in respect to the impact on the students’ time
management and autonomy development. Addition-
ally, it is also important to understand the definition
of virtual assistants according to the authors and how
are they implemented.
This literature review follows the approach sug-
gested by Materla et al. (Materla et al., 2017) and by
Subhash and Cudney (Subhash and Cudney, 2018). It
is composed of three phases, starting with the plan-
ning, followed by the operation (conducting) and dis-
semination (reporting) phases.
The planning phase included both the definition
of the bibliography databases and the selection of
the query term. The selected databases were Sco-
pus and Web of Knowledge (WoK), since they pro-
vide the most relevant and accurate results. The
query used the term (\virtual assistant" AND
\higher education") for the title, keywords an ab-
stract. The papers were further restricted to the last
20 years, from January 1st, 2000 and December 31st,
2019. Only papers available in the institutional repos-
itories, peer reviewed and written in English were
considered.
The second phase (conducting) started with the se-
lection of the relevant papers and exclusion of the re-
maining. Repeated papers and papers without avail-
able full-text PDF were excluded. After these step,
a total of 51 papers remained. The process contin-
ued with the selection of the most relevant articles,
through title and abstract analysis. Non relevant pa-
pers were eliminated. Subsequently, the most relevant
papers remained, in a total 18 (Table 1).
Table 1: Number of papers during the analysis.
Source Search
results
Repeated and
full-text
title and
abstract
Scopus 135 32 13
WoK 115 19 4
Total 250 51 17
The third phase consists of the reporting of the re-
sults.
3 ANALYSIS AND DISCUSSION
The analysis started with the characterization of the
final 17 papers. A content analysis followed, to as-
sess the context and definition of virtual assistants in
education, according to the purpose of this work.
In total, papers from 13 countries were found.
England was the country with the highest number of
papers (3). Russia and Switzerland followed, with 2
papers each. The remaining countries, namely Singa-
pore, Pakistan, Taiwan, Canada, India, France, Bul-
garia, Saudi Arabia, Spain and Germany are also
mentioned with 1 paper each (Figure 1).
Canada
India
Switzerland
France
Bulgaria
Saudi
Arabia
Spain
Germany
Ta iw an
England
Russia
Pakistan
Singapore
Figure 1: Number of papers per country.
It seems that the number of papers per year has
steadily been growing since 2000, with the exception
CSEDU 2020 - 12th International Conference on Computer Supported Education
98
of 2010 (Figure 2). This indicates that the field is still
new and researchers are starting to invest time and ef-
fort in the development of this area.
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Figure 2: Papers published each year.
Most of the studies originate in the technological
area, conducted by scientists from Computer Science,
Informatics and Information technologies. Some also
fall under the Education and Learning Technologies
(Figure 3).
Computer
Science
Informatics
and
Engineering
Education and
Learning
Technology
Figure 3: Papers published in each area.
After the characterization, the papers content was
analysed with the assistance of text mining algorithms
and techniques.
3.1 Text Mining
The PDF files were used to build a text corpus. All
the process was performed in R, using the tidytext
package (https://github.com/juliasilge/tidytext). The
corpus was pre-processed, removing repeated forms,
building a dictionary of terms, eliminating irrelevant
words and minimizing the number of different words
through reduction of inflectional form of the words
(stemming).
After this initial step, a histogram of unigrams
(terms composed of a single word) and bigrams
(terms with two consecutive words) were calculated
(Figures 4 and 5).
assist
perform
environ
learner
question
develop
process
support
inform
educ
intellig
model
user
base
agent
data
system
learn
student
0 200 400 600
Figure 4: Unigram frequency for the whole corpus.
artificial intelligence
data mining
learning environment
learning environments
learning path
machine learning
natural language
subject matter
swarm intelligence
virtual assistant
0 10 20 30 40
Figure 5: Bigram frequency for the whole corpus.
As expected, both figures suggests that the main
focus of the papers is related to technology, with
application in education and learning. Most of the
words and bi-words refer to student, machine and ar-
tificial intelligence, learning environments, virtual as-
sistant, and others. The focus is, clearly, on education,
supported by intelligent, virtual assistants, as we ex-
pected from the search previously conducted. “Vir-
tual assistant” is the most frequent keyword which
was founded in the papers. The second important key
phrase is “intelligent learning systems”. The intelli-
gent learning system means the learning management
system with an embedded virtual assistant, tutor, chat-
bot or intelligent agent. A really common case when
creating a whole system, except for a virtual assis-
tant. Also, frequent concept is a virtual assistant for
learning analytics. It means that virtual assistant is an
application or web-service which helps students to get
their learning analytics. One of the most popular con-
cepts of a virtual assistant is a chatbot. The idea is so
admired because there is a desirable opinion that the
most convenient way of the student interconnection
and learning environment is a voice assistant. There
are plenty of studies that create a chatbot. We only
take two the most interesting researches (cases). Ar-
tificial intelligence is applied more often.
After the global frequency of words and bi-words,
Virtual Assistants for Learning: A Systematic Literature Review
99
a comparative relevance analysis was made, through
the Term Frequency/Inverse Document Frequency
(TF-IDF) numerical statistic index. This numerical
statistic reflects how important a word is to a docu-
ment in a collection of papers. In other words, it as-
sesses the most frequent terms and the most relevant
terms for each document (Figure 6).
From these figures it is possible to understand
what each paper values more, in comparison with the
others. This gives a clue about the main focus of each
paper, which provides a base for the category identi-
fication through a qualitative analysis.
3.2 Content Analysis
The main research question is to understand the influ-
ence of virtual assistants on students autonomy and
motivation. The resulting papers approach this in sev-
eral ways. Albalowi and Alhamed (Albalowi and Al-
hamed, 2017) start by categorizing the most influenc-
ing factors for student retention:
1. Academic Integration.
2. Social Integration.
3. Institutional Commitment.
4. Out-of-institution factors.
They conclude that virtual assistants could be ap-
plied in the first three factors, thus providing sup-
port for students in this issues and helping prevent-
ing retention. The first area is academic integration,
where learning analytics through machine learning
algorithms can help preventing students’ difficulties.
They use big data and multiple data types to exam-
ine students’ performance. Two predictive models are
created: the first model was built using the traditional
structured data to predict the students’ final perfor-
mance, and the second was built using the same data
in addition to the students’ sentiment scores. The stu-
dents’ sentiment scores were calculated by analyzing
their textual feedback using the Stanford Core NLP
Natural Language Processing Toolkit. They can sup-
port students by providing them learning analytic and
failure prediction.
This is further supported by Ciolacu et al. (Cio-
lacu et al., 2018). The authors use Learning Analyt-
ics (LA) for an Early Recognition System with Ma-
chine Learning for a personalized email for prompt-
ing the students of risk. They applied machine
learning algorithms such as Support Vector Machine
(SVM) and Neural Networks (NN) to reduce the fail-
ure rate in examinations. To motivate, especially first
year students, activity diagram of former students are
shown. This underlines the activities in the mathe-
matics course during the semester and grades/scores
(high scores of those who pass the exam).
The second area is social integration. Lamontagne
et al. (Lamontagne et al., 2014) affirm that it is one
of the most substantial part of students’ success. Co-
incidentally, it is the most difficult problem to solve.
There is no easy way to create chatbots or computer
algorithm that can replace real people communica-
tion, simulate relationship in group, discuss and ac-
tually foster collaboration between students. Some-
times, social integration plays an integral role of the
successful education. They even define “a virtual as-
sistant as mainly an information service”.
The third area is an institutional commitment.
There are a lot of papers which consider especially
this area: chatbot, intelligent agents, tutors, and oth-
ers. The AICMS with Dialog Flow is a system that
runs as a messenger end point in Facebook, which
takes the text and voice as input and provides answers
as text and voice as well. It provides all required in-
formation about the college, student information, ex-
amination section information, placement cell infor-
mation, and cultural events (Arun et al., 2019).
Laeeq and Memon (Laeeq and Memon, 2019)
presented their “Scavenge” parencite. The Scav-
enge is an AI-based Intelligent Multi-Agent System
(MAS), developed using JADE (Java Agent Develop-
ment Framework). The authors proposed their vari-
ant of the system as three agents platform. The con-
cept was proposed earlier by Zhang et al. (Zhang
et al., 2010), where the agents assume the roles of
the Learner, the Teacher and the Platform, activelly
participating in the educational process (Zhang et al.,
2010).
Zakharova (Zakharova, 2018) has created a pro-
fessional development forecast model. She used ma-
chine learning algorithms to analyze students’ papers.
Pre-processing and statistical analysis of texts were
carried out to highlight features characterizing the
general vocabulary and the use of general scientific
and professional terminology. These approaches al-
lows students to get immediate feedback, contribut-
ing to better learning and academic success. In fact,
this literature review further confirms that the effec-
tiveness of a virtual assistant depends on the quality
of feedback.
There are three kinds of feedback elements given
by the virtual assistant: evaluation, disturbances and
hints. Disturbances are questions that are intended to
test the confidence of the students about their work.
Hints are documents (text, graphs, formulas, etc.) re-
lated to a particular question of the experimental pro-
tocol that help students find by themselves possible
errors or problems (Geoffroy et al., 2002).
CSEDU 2020 - 12th International Conference on Computer Supported Education
100
rodriguez_2008.pdf
tran_2009.pdf
wong_2010.pdf
zakharova_2018.pdf
zhang_2010.pdf
laeeq_2019.pdf
lamontagne_2014.pdf
lin_2012.pdf
nenkov_2016.pdf
pedrazzoli_2009.pdf
pogorskiy_2019.pdf
albalowi_2017.pdf
arun_2019.pdf
ciolacu_2018.pdf
currie_2016.pdf
geoffroy_2002.pdf
khasianov_2017.pdf
Figure 6: Unigram TF-IDF index.
Pedrazzoli and dall’Acqua (Pedrazzoli and
dall’Acqua, 2009) proposed a concept of a platform
that recognizes three general levels of advice: a) a re-
minder of the current target; b) a general description
of how to achieve the solution; c) a description of
exactly which problem solving action steps should be
taken. Each of these three levels may be represented
by multiple assistance steps. They assume that a
personal assistant should be able to: recognize a
large variety of student solutions; diagnose student
“Subject Matter” understanding and recommend
target oriented, optimized “learning approach adap-
tations”; tailor tutorial actions accordingly; support
collaboration; support specific forms of adaptation
for collaboration activities, like recommending
suitable collaborators and actions; adapt the interface
to facilitate collaboration activities (enforce specific
roles and rules); advise students how to interact
efficiently; reasoning, specify techniques to acquire
and propose additional knowledge material about a
domain or subject matter; use the knowledge base
to solve problems in that domain or subject matter;
support educational workflow sequences.
In addition to supporting the workflow of infor-
mation, virtual assistants can also be used as tools for
time management and motivation. Each student has
a specific profile, with different academic, social and
economic backgrounds. Each student will have his
own time, effort and process for learning. A teaching-
learning process that is too much different and incom-
patible with the student’s profile can lead to dropout.
Thus, dropout prevention can be approached with per-
sonalized learning environments, to best fit each stu-
dent’s abilities and restrictions. Tran et al. (Tran et al.,
2009) suggests a personal tutor, an intelligent assis-
tant that follows students during their learning and
helps students to overcome their difficulties. If nec-
essary, it can refer the student to other support mech-
anisms, such as counsellors, placement officers, ex-
amination officers, accommodation officers etc., each
of whom can assume the support role in their specific
context.
There is a dependency between the implementa-
tion of virtual assistant and student learning motiva-
tion. Surprisingly, students had almost no interest in
using contact via phone or social media. According
to the survey, students currently prefer communicat-
ing with tutors face-to-face and by email (Currie et al.,
2016). According to these authors, the virtual assis-
tant should be implemented as a voice service, 3D
agent or avatar, to foster motivation. They are more
user-friendly, handy and generally more interesting
for students. The results of the study indicated that
“students expressed positive interest in avatars and
their motivation to learn”.
A classical representation of virtual assistant con-
sists the roles of teacher and encourager. In one
step further, establishing the three-agent learning as
the learner-system-teacher, the system might be more
than just a ”mentor”. According to the study expe-
rience of using three part system shows that the en-
gagement and the involvement of the students that the
immediate proper feedback and problem-based learn-
ing provide give measurable and profound advantage
over the ordinary education process (Khasianov et al.,
2017).
Lin et al. (Lin et al., 2012) focus on the under-
standing of student’s states. The system is based on
dual-mode operation: facial expression recognition,
and text semantics as the main elements in affective
computing to understand users’ emotions. Text se-
mantics are used to understand learner’s learning sta-
tus, and the results contribute to course management
agents in order to choose the most appropriate teach-
ing strategies and feedback to the users. According to
the authors, emotions is an important factor for learn-
Virtual Assistants for Learning: A Systematic Literature Review
101
ers’ motivation, whereas motivation plays a vital role
in knowledge development.The use of emotion recog-
nition in the learning system had good feedback from
students. According to the study there was an increase
in user interest in learning digital art.
Nenkov et al. (Nenkov et al., 2016) used arti-
ficial intelligent technologies to simplify scheme of
academic interaction. The idea is to use special AI
agent as third part in Facebook messenger. The in-
teraction between the teacher and the student is car-
ried out in the online environment (in social net, for
example) through mediation of an AI agent (chatbot)
which acts on the basis of knowledge of the LMS ser-
vices. This AI agent is likely to show competencies in
a narrow area (which is limited by the content of the
course topic) in combination with the general commu-
nication and intelligence skills. The feature enables
for the students to integrate learning in familiar social
networks and for the teachers to reduce the burden
and relieve them from monotonous work.
Rodriguez et al. (Rodriguez et al., 2008) use
AI technologies in a couple of chatbot: T-BOT and
Q-BOT. These are two virtual assistants designed to
tutor and evaluate students into an e-learning plat-
form. Both have been designed as PHP modules so
they can be easily integrated into e-learning platforms
like Moodle. The Q-BOT and T-BOT entities gives
students opportunity to evaluate and monitor their
progress as a real one, guiding and tutoring students
in their access to platform’s resources, T-BOT is able
to answer students’ questions about different subjects
using natural language. They are both great instru-
ments integrated into e-learning system, designed to
create friendly and clear learning environment. It
should be noted that this is an essential requirement
for such systems, where the main factor influencing
student motivation is convenience and comprehensi-
bility.
Another requirement is an adaptive and person-
alised assistant. Adaptive navigation helps students
locate and navigate in information hyperspace. Adap-
tive presentation adapts the contents or display of a
page according to the user’s profile. Adaptive collab-
oration helps learners to find the most suitable helpers
or collaborators (Wong and Looi, 2010). To meet
these requirements the authors used swarm intelli-
gence, a set of computing algorithms in the form of
multi-agent systems that simulate how swarms of in-
sects or birds move or work. The key feature of swarm
intelligence is “to balance exploitation (of the known
best solution) and exploration (of the unknown solu-
tions)”. This feature in such swarm systems facilitates
adaptivity in dynamic or unstable environment. Ac-
cording to the authors, swarm intelligence has proven
efficient in solving highly complex optimisation prob-
lems or pattern recognition and plays a key role to im-
prove adaptivity of educational systems.
Pogorskiy et al. (Pogorskiy et al., 2019) discuss
the design of a virtual learning assistant which con-
sists of an extension to the Chrome web browser to
be used as an instrument to support self-regulated
learning. The extension allows for the collection of
data on its users’ web sessions, and interacts with
the user through pop-up notifications and the ex-
tension’s dashboard located in the Chrome menu.
Micro-randomized notification can be useful for self-
regulation.
4 CONCLUSIONS
In this paper we have explored the research in the field
of virtual assistants in higher education. The review
provides an array of uses and techniques for devel-
oping virtual assistants and how they can be used for
education support.
Student support can be categorized depending on
the role of virtual assistant in four types: digital tutor,
the digital secretary, the motivator agent, the mentor
agent. The tutor’s role is to help students through
the learning process The digital secretary helps or-
ganising the learning process, solving administrative
issues, and remind deadlines. The motivator helps
integrating in social environments, to help overcome
stress and anxiety, remind the main aim and motivate
students. Finally, the mentor gives a general descrip-
tion of how to achieve the solution and description of
exactly which problem solving action steps should be
taken.
Virtual assistants are becoming popular and useful
technology, with a variety of advantages, contributing
to automation of tasks and providing support for stu-
dents in time-management, access to information and
communication facilitation. The technology is still in
its infancy. There are many aspects that are neces-
sary to improve to make virtual assistant effective in
student motivation and engagement.
REFERENCES
Albalowi, A. and Alhamed, A. (2017). Big data and learn-
ing analytics in higher education: Demystifying vari-
ety, acquisition, storage, NLP and analytics.
Arun, K., Nagesh, A., and Ganga, P. (2019). A Multi-
Model and Ai-Based Collegebot Management System
(Aicms) for Professional Engineering Colleges. In-
ternational Journal of Innovative Technology and Ex-
ploring Engineering, 8(9):2910–2914.
CSEDU 2020 - 12th International Conference on Computer Supported Education
102
Ciolacu, M., Tehrani, A. F., Binder, L., and Svasta, P. M.
(2018). Education 4.0 - Artificial Intelligence Assisted
Higher Education: Early recognition System with Ma-
chine Learning to support Students’ Success. In 2018
IEEE 24th International Symposium for Design and
Technology in Electronic Packaging (SIITME), pages
23–30, Iasi. IEEE.
Correia, A. M. R. and Mesquita, A. (2006). Novos p
´
ublicos
no ensino superior desafios da sociedade do conheci-
mento. Edic¸
˜
oes S
´
ılabo, Lisboa. OCLC: 494458369.
Currie, E., Harvey, P. H., Daryanani, P., Augusto, J. C., Arif,
R., and Ali, A. (2016). An investigation into the effi-
cacy of avatar-based systems for student advice. EAI
Endorsed Transactions on e-Learning, 3(11):151450.
Geoffroy, F., Aimeur, E., and Gillet, D. (2002). A virtual as-
sistant for web-based training in engineering educa-
tion, volume 2363 of Lecture Notes in Computer Sci-
ence (including subseries Lecture Notes in Artificial
Intelligence and Lecture Notes in Bioinformatics).
Hanus, M. D. and Fox, J. (2015). Assessing the effects of
gamification in the classroom: A longitudinal study
on intrinsic motivation, social comparison, satisfac-
tion, effort, and academic performance. Computers
& Education, 80:152–161.
Khasianov, A., Suleymanov, D., and Marchenko, A. (2017).
Three agent platform approach for digital education
environment. pages 8580–8587, Valencia, Spain.
Laeeq, K. and Memon, Z. (2019). Scavenge: an intelligent
multi-agent based voice-enabled virtual assistant for
LMS. Interactive Learning Environments.
Lamontagne, L., Laviolette, F., Khoury, R., and Bergeron-
Guyard, A. (2014). A framework for building adap-
tive intelligent virtual assistants. Proceedings of the
IASTED International Conference on Artificial Intel-
ligence and Applications, AIA 2014, pages 342–349.
Lin, H.-C. K., Wang, C.-H., Chao, C.-J., and Chien, M.-
K. (2012). Employing Textual and Facial Emotion
Recognition to Design an Affective Tutoring System.
Turkish Online Journal of Educational Technology,
11(4):418–426. WOS:000309531600042.
Lopes, R. P., Mesquita, C., de G
´
ois, L. A., and dos San-
tos J
´
unior, G. (2019). Students’ Learning Autonomy:
A Systematic Literature Review. pages 5958–5964,
Palma, Spain.
Lopes, R. P. and Mesquita-Pires, C. (2013). Novos p
´
ublicos
no Ensino Superior Polit
´
ecnico: desafios de formac¸
˜
ao.
In Membiela, P., Casado, N., and Cebreiros, M. I.,
editors, III International Conference on Higher Ed-
ucation Teaching, pages 535–539. Educaci
´
on Editora,
Ourense, Spain.
Materla, T., Cudney, E. A., and Antony, J. (2017). The ap-
plication of Kano model in the healthcare industry: a
systematic literature review. Total Quality Manage-
ment & Business Excellence, 0(0):1–22.
Mesquita, C., Lopes, R. P., Garc
´
ıa, J. l., and de la Cruz del
R
´
ıo Rama, M. (2015). First Contact with the Word of
Work: The Competence Built in the Teaching Prac-
tices. In Peris-Ortiz, M. and Merig
´
o Lindahl, J. M.,
editors, Sustainable Learning in Higher Education SE
- 6, Innovation, Technology, and Knowledge Manage-
ment, pages 75–87. Springer International Publishing.
Mesquita, C., Lopes, R. P., Garc
´
ıa, J. l., and Rama, M. d.
l. C. d. R. (2014). Pedagogical Innovation in Higher
Education: Teachers’ Perceptions. In Peris-Ortiz, M.,
Garrig
´
os-Sim
´
on, F. J., and Gil Pechu
´
an, I., editors,
Innovation and Teaching Technologies: New Direc-
tions in Research, Practice and Policy, pages 51–60.
Springer International Publishing, Cham.
Nenkov, N., Dimitrov, G., Dyachenko, Y., and Koeva, K.
(2016). Artificial intelligence technologies for person-
nel learning management systems. pages 189–194.
Pedrazzoli, A. and dall’Acqua, L. (2009). OPUS One An
Artificial Intelligence-Multi Agent based Intelligent
Adaptive Learning Environment ( IALE ).
Pogorskiy, E., Beckmann, J., Joksimovic, S., Kovanovic,
V., and West, R. (2019). Utilising a Virtual Learning
Assistant as a Measurement and Intervention Tool for
Self-Regulation in Learning. pages 846–849.
Rodriguez, E., Burguillo, J., Rodriguez, D., Mikic Fonte,
F., Gonz
´
alez Moreno, J. C., and Novegil, J. (2008).
Developing virtual teaching assistants for open e-
learning platforms. pages 193–198.
Rosado-Pinto, P. (2008). Formac¸
˜
ao pedag
´
ogica no ensino
superior.: O caso dos docentes m
´
edicos. Revista de
Ci
ˆ
encias da Educac¸
˜
ao, (7):111–124.
Sim
˜
ao, J. V., Costa, A. d. A., and Santos, S. M. d. (2002).
Ensino superior: uma vis
˜
ao para a pr
´
oxima d
´
ecada.
Gradiva, Lisboa. OCLC: 917200006.
Subhash, S. and Cudney, E. A. (2018). Gamified learning
in higher education: A systematic review of the litera-
ture. Computers in Human Behavior, 87:192–206.
Tenorio, M., Reinaldo, F., Esperandim, R., Lopes, R., Gois,
L., and Dos Santos Junior, G. (2018). C
´
eos: A collab-
orative web-based application for improving teaching-
learning strategies. In Advances in Intelligent Systems
and Computing, volume 725, pages 107–114.
Tran, T. D., Bouchon-Meunier, B., Marsala, C., and Putois,
G.-M. (2009). An Intelligent Assistant to Support Stu-
dents and to Prevent them from Dropout. In CSEDU.
Veiga Sim
˜
ao, A. M., Flores, M. A., Fernandes, S., and
Figueira, C. (2008). Tutoria no ensino superior:
concepc¸
˜
oes e pr
´
aticas. S
´
ısifo/Revista de Ci
ˆ
encias da
Educac¸
˜
ao, (7):15.
Wong, L.-H. and Looi, C.-K. (2010). Swarm intelligence:
New techniques for adaptive systems to provide learn-
ing support. Interactive Learning Environments, 20.
Zakharova, I. (2018). Machine Learning Methods of Pro-
viding Informational Management Support for Stu-
dents’ Professional Development. The Education and
science journal, 20:91–114.
Zhang, Y., Oussena, S., Clark, T., and Kim, H. (2010). Use
Data Mining to Improve Student Retention in Higher
Education - A Case Study. In Proceedings of the
12th International Conference on Enterprise Informa-
tion Systems, pages 190–197, Funchal, Madeira, Por-
tugal. SciTePress - Science and and Technology Pub-
lications.
Virtual Assistants for Learning: A Systematic Literature Review
103