Social Robots as Teaching Assistance System in Higher Education:

Conceptual Framework for the Development of Use Cases

Josef Guggemos, Michael Burkhard, Sabine Seufert and Stefan Sonderegger

Institute of Business Education and Educational Management – Digital Learning, University of St. Gallen,

St. Jakobstrasse 21, 9000 St. Gallen, Switzerland

Keywords: Social Robots, Humanoid Robots, Human Robot Interaction, Symbiotic Design, Teaching Assistance System.

Abstract: This paper provides an overview of the current state of research on social robots in higher education and the

existing frameworks to categorize and develop social robot applications. Based on the existing work, we

present our own framework to develop use cases for social robots in the education sector. Our framework is

based on a heuristic and symbiotic design approach that serves as a guideline for developing use cases and

views human-robot interaction as two complementary and mutually reinforcing roles. We illustrate our

framework by means of a use case that we have conducted in 2019 during the initial lecture of the large-scale

course ‘Introduction to academic writing’.

1 INTRODUCTION

Higher education faces a highly dynamic

environment. In light of the current technological

developments, an extensive substitution of human

labor by smart machines (artificial intelligence) may

come to the fore (King & Grudin, 2016; Nedelkoska

& Quintini, 2018). In this context, Davenport and

Kirby (2016) put the focus on mutual

complementation and collaboration (augmentation),

i.e., “people and computers supported each other in

the fulfilment of tasks” (p. 2). According to Jarrahi

(2018), augmentation can be conceptualized as a

“Human-AI symbiosis” where the collaboration

between humans and artificial intelligence (AI) can

make both parties smarter over time (p. 583). This

kind of symbiosis may change if the communication

partner takes on a physical form through social

robots. Due to the increasing attention to AI and

Human-Computer-Interactions or more specific

Human-Robot-Interactions (HRI), the development

and use of AI-based robots is recently an emerging

field in many areas such as medicine, finance, service

industries, and education (Thimm et al., 2019).

Social robots increasingly pervade the daily life.

Breazeal (2003) refers to social robots as machines

“that people apply a social model to in order to

interact with and to understand” (p. 167).

Social robots have the potential to become

integral part of the educational infrastructure (Mubin,

Stevens, Shahid, Al Mahmud, & Dong, 2013;

Belpaeme, Kennedy, Ramachandran, Scassellati, &

Tanaka, 2018). However, until now, the

implementation of social robots in education has been

rather scarce since it is a relatively new emerging

research field and requires considerable resources.

Instead, studies have often tried to implement

pedagogical agents and traditional intelligent tutoring

systems in learning scenarios (e.g., Baker, 2014). In

contrast to these learning technologies, social robots

interact with students in a synchronous way making it

possible to react on individual intents with a physical

presence. The potential of social robots as assistant

systems in education is a rather new phenomenon. In

a recent literature review, Belpaeme et al. (2018)

summarized the state of the art. Assuming that social

robots increasingly pervade future workplaces,

students may need training to efficiently collaborate

with digital assistants. Since digital assistants might

be social robots in the future, students should be able

to understand the future today and acquire skills to

help shaping the future in terms of using social robots

with focus on augmentation and a fruitful symbiotic

approach.

Guggemos, J., Burkhard, M., Seufert, S. and Sonderegger, S.

Social Robots as Teaching Assistance System in Higher Education: Conceptual Framework for the Development of Use Cases.

DOI: 10.5220/0009794801250132

In Proceedings of the 12th International Conference on Computer Supported Education (CSEDU 2020) - Volume 1, pages 125-132

ISBN: 978-989-758-417-6

125

2 DEFINITION OF SOCIAL

ROBOTS

Due to the wide variety of different appearances of

social robots, it is necessary to classify them. Table 1

elaborates the characteristics of social robots by

distinguishing between technical and social

dimensions. The first technical dimension to classify

a social robot is the physical presence and thus the

design of the robot from abstract to human-like

(Baraka, Alves-Oliveira, & Ribeiro, 2019). The

second technical dimension comprises the degree of

autonomy of the robot, ranging from remote-

controlled to completely autonomous. The social

dimensions (Breazeal, 2003) show on the one hand

the development stage of the robot interaction model

and on the other hand the social embedding in the

environment. The interaction can range from

evocative and rather passive to sociable and

proactive. The dimension of social embedding adds a

broader focus on the social behavior and integration

into the environment. It ranges from pure perception

and reaction to the social environment to a socially

intelligent robot with full social competence (Fong,

Nourbakhsh, & Dautenhahn, 2003).

3 RESEARCH GOALS AND

METHODS

The aim of the paper at hand is to investigate the

potential of social robots for educational purposes in

higher education because there is a research gap in

terms of pedagogical uses and the robots’ social

capabilities. In this vein, it may be important to

investigate whether the social robot can be useful in

the social environment as an autonomous system and

how the interaction between human and robot is

changing over time. Conceptual frameworks might be

useful for designing use cases as an iterative pilot

testing. Hence, the paper at hand might act as a

stepping stone for coming researchers who might

more eﬃciently uncover further potential of the

technology, e.g., type of robot to use, how to adapt it

properly to a use case, what kind of architecture the

robot system might need, how to achieve the greatest

pedagogical value, etc.

In light of the identified research gap, the

following overarching research question should be

addressed:

How can use cases be designed for social robots as

assistance systems in higher education to improve the

learning process and enhance learning experiences

(e.g., reaching new learning goals) of higher-

education students?

The objectives of the paper at hand are therefore

twofold:

 Analysis of empirical studies with social robots

in order to investigate underlying assumptions,

goals, methods and empirical results for

designing and evaluating the use cases;

 Development of a conceptual framework as an

appropriate methodology to theoretically

founded develop use cases for social robots as

assistant systems in higher education.

To this end, we lay the foundation for our framework

in section 4 by conducting two literature reviews.

First, we look at how social robots have been used in

higher education. Second, we provide an overview of

Table 1: Characteristics of social robots as socio-technical systems.

Note: Draws on the work of Breazeal (2003), Duffy (2003), Fong et al. (2003), Belpaeme et al. (2018), and Baraka et al.

(2019).

CSEDU 2020 - 12th International Conference on Computer Supported Education

126

existing frameworks in the field of social robots.

Section 5 lays out our own extended framework.

Section 6 concludes with some final remarks.

4 LITERATURE REVIEW

4.1 Context: Social Robots in Higher

Education

The EBSCOhost database and the IBM Science

Summarizer Beta database were searched to find

relevant literature focusing on the use of social robots

in higher education. The abstracts were searched for

terms such as humanoid robots or social robots or

higher education and university or college or lecture

or post-secondary or postsecondary.

The search procedure yielded a total of 20

relevant papers. The earliest study appeared in 2012,

the latest study was published in 2019. Four

contributions were literature reviews. A majority of

fourteen contributions analyzed the deployment of

humanoids in lectures to foster students’ learning

outcome. The remaining two studies examined

specific aspects of the topic, e.g., design of

humanoids in higher education.

Clustering the studies according to their subjects

revealed that eight studies focused on STEM, one

study focused on Business, one study focused on

STEM & Business, one study focused on Languages,

and five studies did not provide information about the

subject.

Four studies focused on undergraduate students,

two studies focused on graduates, one study focused

on graduates and undergraduates and nine studies did

not provide information about the university level.

In terms of the educational setting, the studies

differed in the following ways: Eight studies were

carried out in a lecture or classroom settings, three in

workshops or as part of a group work, two in a

laboratory environment and three studies did not

provide the necessary information.

The roles that the social robots took in the studies

also varied. In five studies the robot acted as a lecturer

or tutor, in two studies as a teaching assistant, in two

studies as a mediator or partner and in three studies

the robot was used as a test platform for the

development of applications. One study used the

robot as an educational means to teach technology

related content (Flynn, 2017). Three studies did not

provide the necessary information.

4.2 Design: Frameworks for Social

Robot Use Cases

The EBSCO database, the IBM Science Summarizer

Beta database and Google Scholar were searched to

find journal articles and conference papers focusing

on conceptual frameworks in combination with HRI.

Search terms were conceptual framework or

theoretical framework or reference architecture or use

Figure 1: The dimensions of the engagement profile (Cooney & Leister, 2019, p.8).

Social Robots as Teaching Assistance System in Higher Education: Conceptual Framework for the Development of Use Cases

127

case and human robot interaction or social robot

interaction. Subsequently, the results were manually

filtered. Overall, fourteen relevant papers were

identified. The earliest study appeared in 2002,

whereas the latest study was published in 2020. We

split them up into three types.

The first type comprises studies that provide

frameworks with robots in automatization processes

and industry related contexts. Studies, which regard

robots more as a tool than as a social counterpart also

belong to this category. Often, they focus on the

technical implementation. For examples, see

Radanliev, Roure, Nicolescu and Huth (2019) or

Cuevas, Fiore, Caldwell and Strater et al. (2007).

The second type consists of studies with

frameworks about social robots in their environment.

Breazeal et al. (2004) propose an early framework

towards robots as partners rather than robots as tools.

You and Robert (2018) provide a framework for

human-robot teamwork. Their framework describes

which characteristics are brought into a work process

by which parties (humans, robots) and how this leads

to team output.

Belanche, Casaló, Flavián and Schepers (2020)

create a theoretical framework for the implementation

of service robots. In three categories (robot design,

customer features, services encounter characteristics)

they identify important factors for a successfull

design and implementation of service robots.

Baraka et al. (2019) provide an extended

framework for characterizing social robots. Their

framework covers along seven dimensions the

interaction and the relational role between robot,

human and the context. In addition, they outline

different approaches for designing human robots:

human-centered design, robot-centered design, and

symbiotic design (Baraka et al., 2019, pp. 31–33). To

develop social robots in a symbiotic design, Baraka

et. al. (2019, p. 33) recommend identifying the

relative strengths and weaknesses of each party. They

refer to the study of Veloso, Biswas, Coltin, and

Rosenthal (2015), in which autonomous robots ask

humans for help with certain activities, such as

pressing the elevator button for them. This little

assistance from humans allows the robots to navigate

on several floors without the need for any robot hands

and makes the implementation of use cases easier and

cheaper.

The third type comprises studies with frameworks

about social robots in the context of education. Yang

and Zhang (2019) develop design guidelines for an

intelligent tutoring robot in the tension field between

human tutor, student, curriculum, and social milieu.

Its scope is relatively narrow, as it only covers the use

case of the tutor and no other potential applications in

the education sector. Cooney and Leister (2019)

provide a more general framework by adapting the

engagement profile to the educational context. In an

exploratory study at a graduate school, they defined

potential useful capabilities to create a prototype for

a robotic teaching assistant. Based on this, they

weekly tested the robot in a classroom and used the

engagement profile to iteratively improve their robot.

Seven contributions were related to type 1, i.e.,

they provide frameworks for robots from a more

technical view in an industry related setting. Five

contributions are type 2 studies with a focus on the

social interaction between humans and social robots.

The remaining two studies were type 3 studies, i.e.,

they provide frameworks for social robots in

education.

5 RESULT: CONCEPTUAL

FRAMEWORK

5.1 Structure of the Framework

Based on the available frameworks, summarized in

the previous section, this chapter lays out our own

extended framework (see Figure 2). Similar to the

framework of Baraka et al. (2019) our framework

focuses on the overall system behavior of robot,

human, and context. While Baraka et al. (2019, p. 3)

define the context as “Purpose and application area”,

we take a broader view of this notion and add further

elements to it.

Knowledge and attitude of the stakeholders

towards social robots may be of central importance

for the successful implementation of a use case. This

applies to both the development team and the users.

To represent the requirements and features of the

users, we utilize the term customer features from

Belanche et al. (2020) who have chosen this element

as a key part of their framework.

Belanche et al. (2020) deal in their framework

also with technical aspects in the form of robot

design. Therefore, we integrate the element

technology into our model. Together with financial,

legal and ethical constraints, they complete the

category context.

Baraka et al. (2019) as well as Cooney and Leister

(2019) define the role of social robots by its

capabilities. In addition, the role of humans may also

be defined by their capabilities. The specific

capabilities of both parties (humans and robots) are

considered to form the use case. We propose that as

CSEDU 2020 - 12th International Conference on Computer Supported Education

128

an evaluation tool the engagement profile of Cooney

and Leister (2019) or various outcome measures (You

& Robert, 2018) could be used.

In contrast to the frameworks that we have drawn

on, we put educational aspects into focus and at the

same time try to keep the model as generic as

possible, to cover a broad variety of use cases.

We consider our framework as a design-

specification tool to serve as a guideline for the

development of own use-cases. In our understanding,

the awareness of the complementary roles of humans

and robots in interaction with the context may help to

avoid pitfalls and to create better use cases.

5.2 Paradigm: Symbiotic Design

We follow the symbiotic design approach as outlined

by Baraka et al. (2019, pp. 31–33). Not every

implementation that is technically possible may also

be useful. It is important to consider the relative

strengths of humans and social robots and create

applications against this background. The purpose of

social robots is not to replace humans, but to support

humans by extending their capabilities where

necessary. Otherwise, social robots will not find

social acceptance in the long run. According to our

understanding, it is important to view the interaction

between humans and robots as two complementary,

mutually reinforcing roles.

5.3 Context

In our framework, we follow a heuristic development

approach that creates solution-oriented applications

for practical use with limited resources. When

developing a use case, the development team faces

multiple restrictions due to the context. As a first step,

the development team should think about the context

and record it in writing. The context later implicitly

defines the scope of action.

Depending on the educational goal and setting, the

project must meet different requirements. The

development team has to meet these requirements and

at the same time deliberately assess their own know-

how and anticipate the know-how and attitude of the

future users.

Note. (1) Baraka et al. (2019), (2) Belanche et al. (2020), (3) Cooney and Leister (2019), (4) You and Robert (2018).

Figure 2: Conceptual framework for the development of use cases with social robots.

Social Robots as Teaching Assistance System in Higher Education: Conceptual Framework for the Development of Use Cases

129

The existing infrastructure, the available

technology and the technology frontier should also be

considered. Against this backdrop, the integration and

the usage of already existing services (e.g., text-to-

speech-services) may be preferable to in-house

development, as in-house developments can be very

expensive or even impossible.

Finally, each development team must stay within

the budget and comply with legal and ethical

restrictions (e.g., data protection policies), which

influence which use-cases can or cannot be

implemented.

In its entirety, the context restricts possible use

cases. At the same time, the context also defines the

scope of action for the role of humans and robots.

5.4 Roles of Robot and Human Being

In our understanding, a use case consists of two

complementary roles: The role of the social robot and

the role of the human being. Both roles should

supplement each other and take into account the

respective strengths. Cooney and Leister (2019)

described potential roles for a social robot. The robot

could take over the role of a tutor outside class, of an

avatar or of a teaching assistant. Many more such

roles for socials robot are possible. In our

understanding, humans also play such a role when

interacting with robots. Depending on the context and

the role of the robot, humans could, for example, take

on the role of a supervisor, a maintainer or a mediator.

5.5 Capabilities

In a second step, the development team should design

the roles of humans and robots and their capabilities.

Each role is defined by at least one capability. Cooney

and Leister (2019) mention reading, greeting,

alerting, remote operation, clarification, and motion

as potential capabilities of a robot teaching assistant.

Depending on the use case, the role and the associated

capabilities will change. To give a second example,

the capabilities of a robot concierge in a museum

could be greeting, reading, informing and orientating.

5.6 Implementation and Evaluation

Measures

Measures for implementation and evaluation are

important as a tool to get feedback and iteratively

improve the design. One possible approach for

implementation might be through the engagement

profile.

The engagement profile was originally used for

installations and exhibits in science centers and

museums (Leister et al., 2017). Cooney and Leister

(2019) adapted the engagement profile to the teaching

case (see Figure 1). They argue that similar to

installations in science centers, a social robot

represents an artefact that the students interact with

during their studies and classes. Along the eight

dimensions of the engagement profile (competition,

narrative elements, interaction, physical activity, user

control, achievements awareness, exploration

possibilities) the capabilities can be defined,

measured on a scale and be reevaluated and adjusted

in an iterative process.

The engagement profile is a promising approach

to measure robot capabilities on scales. However,

further research is needed as the dimensions of the

model come from the world of museums and science

centers and may not always fit into a social robot

setting. In addition, the question arises how the

engagement profile is to be used to evaluate the team

performance of humans and robots together.

At this point, the work of You and Robert (2018)

might offer a viable approach. They distinguish

between different team outputs (taskwork outcomes,

teamwork outcomes, subjective outcomes) of robot-

human teams, which could be measured and

evaluated.

5.7 Example for an Implementation

The following section illustrates our framework by

means of a use case that we conducted in 2019 during

the initial lecture of the large-scale course

‘Introduction to academic writing’ (see Figure 3). The

course was mandatory for all the 1,552 freshmen at

our university who were on average 19.77 years old.

The course has an English track (470 students) and a

German one (1,082 students). Therefore, the lecture

was conducted in both English and German.

Apart from our primary goal of the course to

introduce students to academic writing, our

educational goal with the robot was to offer the

students a representative sample of tasks that a social

robot might perform within the context of learning. In

the lecture we also wanted to inform about plagiarism

and plagiarism software.

For the implementation of our use case we could

count on the support of raumCode, a company

specialized on humanoid robots and artificial

intelligence, as well on the educational know-how

from our chair of Business Education. A type Pepper

model (SoftBank Robotics) owned by our chair is at

our disposal.

CSEDU 2020 - 12th International Conference on Computer Supported Education

130

Figure 3: Use case: Social robot Lexi as a teaching assistant in an academic writing course at university.

Based on the given context, we decided to put the

robot in a co-role with the lecturer. The robot should

support the lecturer during the course as an assistant

and showcase model. In return, the lecturer should

lead the lesson and supervise the action of the robot.

In the sense of a complementary and symbiotic

distribution of roles between humans and robots, we

found it appropriate to let the robot do those parts of

the lecture that are related to computer science in the

broadest sense. In this area, the robot can play its

natural strength, because, unlike humans, it can be

connected to application programming interfaces

(APIs), for example.

Among others, the robot in our course explained

plagiarism and its detection by plagiarism software -

and searched for the lecturer's sources for

"greenwashing" in the database of our university

library.

6 DISCUSSION AND OUTLOOK

By means of our literature reviews, we could identify

several studies that explore the use of social robots in

higher education and beyond. However, a number of

challenges for the use of socio-technical systems must

also be taken into account. The introduction of these

technologies into pedagogical practice involves the

solution of technical challenges and requires changes

in pedagogical practice. Moreover, ethical concerns

have to be addressed (Belpaeme et al., 2018). To what

extent it is desirable to delegate education to social

robots has to be discussed in-depth. In this discussion

criteria that go beyond learning efficiency, i.e.,

learning outcomes and costs, should be considered.

The main contribution of our paper is the

development of a conceptual framework in order to

derive theoretically sound use cases for social robots

as assistance systems in higher education. We have

demonstrated the usefulness of the framework by

illustrating our empirical study with the social robot

Pepper in an academic writing course.

In several development cycles, innovative

practical solutions are to be developed, which at the

same time are to produce theories with saturated

evidence that can be used as research results. The

transferability of the innovation developed is less to

be found in the problem solution itself, but rather in

the development of transferable theories: “Theory

informing practice is at the heart of the approach, and

the creation of design principles and guidelines

enables research outcomes to be transformed into

Social Robots as Teaching Assistance System in Higher Education: Conceptual Framework for the Development of Use Cases

131

educational practice” (Reeves, Herrington, & Oliver,

2005, p. 107).

An open question is still how to evaluate the use

cases and in particular the human-robot interactions.

From a symbiotic research paradigm, the evaluation

should focus on the behavior of both partners, human

and robot: Does the human adapt to the robot, and the

robot adapt to the human, in a way that benefits the

interaction? At the current technological state there is

much room for improvement in terms of the human-

robot relationship. The goal should be that robots act

in a way that could be regarded as social in its original

sense.

REFERENCES

Baker, R.S. (2014). Educational data mining: An advance

for intelligent systems in education. IEEE Intelligent

systems, 29(3), 78–82.

Baraka, K., Alves-Oliveira, P., & Ribeiro, T. (2019). An

extended framework for characterizing social robots.

ArXiv Preprint ArXiv:1907.09873.

Belanche, D., Casaló, L. V., Flavián, C., & Schepers, J.

(2020). Service robot implementation: a theoretical

framework and research agenda. The Service Industries

Journal, 40 (3-4), 203–225.

Belpaeme, T., Kennedy, J., Ramachandran, A., Scassellati,

B., & Tanaka, F. (2018). Social robots for education: a

review. Science Robotics, 3(21), 1–9.

https://doi.org/10.1126/scirobotics.aat5954

Breazeal, C. (2003). Toward sociable robots. Robotics and

Autonomous Systems, 3-4(42), 167 175.

Breazeal, C., Brooks, A., Chilongo, D., Gray, J., Hoffman,

G., Kidd, C., Lee, H., Lieberman, J., & Lockerd, A.

(2004). Working collaboratively with humanoid robots.

In 4th IEEE/RAS International Conference on

Humanoid Robots, 2004. Symposium conducted at the

meeting of IEEE.

Cooney, M., & Leister, W. (2019). Using the engagement

profile to design an engaging robotic teaching assistant

for students. Robotics, 8 (1), 21.

Cuevas, H. M., Fiore, S. M., Caldwell, B. S., & Strater, L.

(2007). Augmenting team cognition in human-

automation teams performing in complex operational

environments. Aviation, Space, and Environmental

Medicine, 78 (5), B63‐B70.

Davenport, T. H., & Kirby, J. (2016). Only humans need

apply: Winners and losers in the age of smart machines

(First edition). New York, NY: Harper Business.

Duffy, B.R. (2003). Anthropomorphism and the social

robot. Robotics and Autonomous Systems, 3-4(42), 177

190.

Flynn, P. (2017). Using an Autonomous Humanoid Robot

as a Pedagogical Platform in the Business Classroom.

Journal of Social Science Studies, 4 (1), 178–188.

Fong, T., Nourbakhsh, I. & Dautenhahn, K. (2003). A

survey of socially interactive robots. Robotics and

Autonomous Systems, 3-4(42), 143 166.

Jarrahi, M. H. (2018). Artificial intelligence and the future

of work: human-ai symbiosis in organizational decision

making. Business Horizons, 61(4), 577–586.

https://doi.org/10.1016/j.bushor.2018.03.007

King, J. L., & Grudin, J. (2016). Will computers put us out

of work? Computer, 49(5), 82–85.

https://doi.org/10.1109/MC.2016.126

Leister, W., Tjøstheim, I., Joryd, G., Brisis, M. de,

Lauritzsen, S., & Reisæter, S. (2017). An evaluation-

driven design process for exhibitions. Multimodal

Technologies and Interaction, 1 (4), 25.

Mubin, O., Stevens, C. J., Shahid, S., Al Mahmud, A., &

Dong, J. J. (2013). A review of the applicability of

robots in education. Journal of Technology in

Education and Learning, 1(209-0015), 13.

Nedelkoska, L., & Quintini, G. (2018). Automation, skills

use and training. OECD Social,

Employment and

Migration Working Papers No. 202. Paris.

https://doi.org/10.1787/2e2f4eea-en

Radanliev, P., Roure, D. de, Nicolescu, R., & Huth, M.

(2019). A reference architecture for integrating the

Industrial Internet of Things in the Industry 4.0. ArXiv

Preprint ArXiv:1903.04369.

Reeves, T., Herrington, J. & Oliver, R. (2005). Design

Research: A Socially Responsible Approach to

Instructional Technology Research in Higher

Education. Journal of Computing in Higher Education,

16(2), 97–116.

Thimm, C., Regier, P., Cheng, I. C., Jo, A., Lippemeier, M.,

Rutkosky, K, Bennewitz, M., Nehls, P. (2019). Die

Maschine als Partner? Verbale und non-verbale

Kommunikation mit einem humanoiden Roboter. In C.

Thimm & T. C. Bächle (Eds.), Die Maschine: Freund

oder Feind: Mensch und Technologie im digitalen

Zeitalter (pp. 109–134). Wiesbaden: Springer.

https://doi.org/10.1007/978-3-658-22954-2_6

Veloso, M. M., Biswas, J., Coltin, B., & Rosenthal, S.

(2015). CoBots: Robust Symbiotic Autonomous

Mobile Service Robots. In IJCAI (p. 4423).

Yang, J., & Zhang, B. (2019). Artificial Intelligence in

Intelligent Tutoring Robots: A Systematic Review and

Design Guidelines. Applied Sciences, 9 (10), 2078.

You, S., & Robert, L. (2018). Teaming up with robots: An

IMOI (inputs-mediators-outputs-inputs) framework of

human-robot teamwork. You, S. And Robert, LP (2017).

Teaming up with Robots: An IMOI (Inputs-Mediators-

Outputs-Inputs) Framework of Human-Robot

Teamwork, International Journal of Robotic

Engineering, (IJRE), 2 (3).

CSEDU 2020 - 12th International Conference on Computer Supported Education

132