Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and

ChatGPT

Dan-Cristian Alb

and Camelia S

erban

Faculty of Mathematics and Computer Science, Babes

Bolyai University, Cluj-Napoca, Romania

Keywords:

Efﬁcient Learning, Virtual Assistants, Cloud Computing, Infrastructure as Code, ChatGPT.

Abstract:

In a society in continuous change due to the astonishing speed with which science and technology develop,

the educational system must keep up with these changes and meet the students with effective learning methods

integrated in intelligent platforms. In this respect, the paper proposed an e-learning platform whose underlying

educational framework is built upon three studied principles that lead to efﬁcient learning, namely: elabora-

tion, retrieval practice and feedback. Furthermore, the core functionalities of the proposed tool consist of a

question proposal system and quiz taking system, elements shaped based on active and collaborative learning.

Nevertheless, the proposed tool’s learning authenticity is ensured by the dedicated voice assistant powered by

Alexa and the integration with ChatGPT, an OpenAI product. As for the validity of the proposed solution, an

ongoing study is in place. The study consists of integrating the proposed tool as part of the didactic activity

for the Data Structures course taken by 1st year students enrolled in the Mathematics and Computer Science

undergraduate program offered by Babes¸-Bolyai University (Cluj-Napoca, Romania).

1 INTRODUCTION

In a dynamic and ever-changing society, learning

methods must adapt to meet the needs of modern

learners. The education system plays an important

role in equipping future generations with the knowl-

edge and skills necessary to navigate an uncertain

world. To accomplish this, educators must utilize ef-

fective teaching approaches that engage and empower

learners, enabling them to actively participate in the

educational process.

It is suggested that education theory should shift

towards a more student-centered approach that em-

phasizes active learning, as supported by various

sources including (Bishop et al., 2014), (Doyle,

2008), (J., 2011), (Rich et al., 2014), (Serban and Ves-

can, 2019), and (Freeman et al., 2014). This approach

plays a crucial role in fostering students’ creativity

and competence in their studies. Essentially, active

learning places the onus of learning on the students

and assigns the teacher the role of facilitator.

The approach outlined above aims to give the

learner a primary role in the learning process and al-

lows them to ”discover” knowledge at their own pace

https://orcid.org/0009-0000-9188-6879

https://orcid.org/0000-0002-5741-2597

or in groups, having a minimal guidance from the

lab instructor, who provides support and encourages

imagination and creativity. This approach, as sug-

gested by sources including (Bishop et al., 2014),

(Doyle, 2008), and (Rich et al., 2014), is also de-

signed to foster the development of communication

and teamwork skills, conﬂict resolution abilities, and

time management competencies that can assist stu-

dents in regulating their learning and achieving their

goals. The student-centered paradigm is based on var-

ious principles of effective learning that have been ro-

bustly supported by research in cognitive and educa-

tional psychology over the past few years.

The current paper proposes to design an e-learning

platform based on three principles of an efﬁcient

learning and to test its efﬁciency on students’ learn-

ing performances in a real teaching context. The plat-

form’s design was based on three principles of effec-

tive learning, namely elaboration, retrieval practice,

and feedback, which have been ﬁrmly supported by

research

The support platform has two core functionalities:

multiple choice question proposal system and quiz

taking system. But the authenticity of the proposed

learning system comes from the following compo-

nents which are integrated in our system: dedicated

voice assistant powered by Alexa and integration with

720

Alb, D. and ÈŸerban, C.

Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and ChatGPT.

DOI: 10.5220/0012147500003538

In Proceedings of the 18th International Conference on Software Technologies (ICSOFT 2023), pages 720-727

ISBN: 978-989-758-665-1; ISSN: 2184-2833

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

ChatGPT, an OpenAI product. In this respect, we

enhance the user experience by allowing interaction

with the platform by means of voice commands and

we acknowledge the power that artiﬁcial intelligence

(shortly AI) powered tools can have in the learning

experience, rather than denying their usage.

As for the validity of the proposed solution, an on-

going study is in place. The study consists of integrat-

ing the proposed tool as part of the didactic activity

for the Data Structures course taken by 1st year stu-

dents enrolled in the Mathematics and Computer Sci-

ence undergraduate program at Babes¸-Bolyai Univer-

sity. Validation of the proposed model will be based

on 2 surveys handed to students and will aim to un-

derstand the student experience of using the proposed

model.

The rest of the paper is structured as follows:

Section 2 builds an overall image over the trends of

research done in the area of E-Learning in the re-

cent years. Section 3 reﬂects the way the proposed

tool abides the principles of an effective learning,

while Section 4 describes the Collaborative learning

strategy integrated by our proposed tool. Section 5

presents the proposed E-Learning software system, its

architecture, the main functionalities and its elements

of authenticity. Section 6 presents our conclusions

and future development possibilities of the proposed

tool.

2 RELATED WORK

The use of computer-based tools to promote active

learning in higher education has gained increasing

attention in recent years, as noted by (Weinstein

et al., 2018b). Various studies, including (Mays

et al., 2020), (Weinstein et al., 2018b), and (Greving

et al., 2020), have examined the efﬁcacy of such tools

in enhancing student engagement and learning out-

comes. This section provides a summary of two re-

lated computer-based tools to facilitate active learning

among students.

Moodle (Devi and Aparna, 2020) is a widely used

open-source learning management system (LMS) de-

signed to facilitate online learning and course man-

agement. It provides a range of features such as

course creation, content management, assessment

tools, and communication tools. It allows to create

and customize course content, track student progress,

and communicate with students in real-time.

Educational institutions of various types, includ-

ing kindergarten through grade 12 (or K-12) schools,

universities, and corporate training programs, utilize

it to provide online and blended learning experiences.

Overall, Moodle is a versatile platform that enables

educators to create and deliver effective online learn-

ing experiences.

Canvas (Fern

andez et al., 2017) is a cloud-based

LMS that provides a user-friendly interface and a

range of course design and customization tools, in-

cluding multimedia support, quizzes and assessments,

and interactive discussions. Canvas also includes

features for tracking and analyzing student perfor-

mance, facilitating communication between educators

and students, and supporting collaborative learning.

Overall, Canvas is a comprehensive LMS that en-

ables educators to create engaging and effective on-

line courses.

Other tools such as Google Meet, Zoom, Mi-

crosoft Teams (Alameri et al., 2020) are also being

used for educational purposes. Zoom and Google

Meet are mainly used for video live streaming, there-

fore enabling online lecture streaming. Whereas, Mi-

crosoft Teams also brings features speciﬁc to LMS

tools such as assignment management, grading man-

agement and even quiz taking through forms. More-

over, Microsoft Teams is taking the game to next

level through its rich available applications integra-

tion within the platform (e.g. Microsoft 365)

In relation to existing approaches, the proposed

tool engages students in active learning and collab-

oration by allowing them to propose multiple choice

questions based on studied class concepts. The ques-

tion design process involves students not only creat-

ing a question, but also providing the rationale for the

answer choices. Teachers and certiﬁed class review-

ers supervise the entire process, and collaborative re-

views are conducted to improve question quality, if

required. Thus, the proposed tool introduces a new

approach to enhance the learning process through ac-

tive learning and collaboration.

3 DESIGNING A TOOL FOR

EFFICIENT LEARNING

Over the past few years, signiﬁcant progress has been

made in the science of learning, enabling us to gain a

better understanding of effective teaching and learn-

ing principles. Through rigorous research, we now

have compelling evidence and speciﬁc recommenda-

tions regarding the strategies that educators and learn-

ers can employ to gain learning efﬁciency, as high-

lighted in (Agarwal and Henry L. Roediger, 2018),

(Dunlosky et al., 2013), and (Weinstein et al., 2018a).

As for the scope of this research activity, three learn-

ing strategies - elaboration, retrieval practice, and

feedback are integrated into the proposed e-learning

Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and ChatGPT

721

platform with the aim of enhancing the student learn-

ing process in their formation as professionals.

3.1 Closer Abides to Elaboration

Principle

Elaboration refers to the process of generating an ex-

planation for why a given fact or concept is true, as

noted in (Dunlosky et al., 2013). This approach helps

connect new information with existing knowledge,

thereby enhancing learning. However, research sug-

gests that it is crucial for students to check their an-

swers with available class materials or teachers when

using this technique, as poor elaboration content may

hinder learning (Clinton et al., 2016).

Our proposed platform, Closer, is used by stu-

dents to elaborate multiple choice questions for as-

sessment tests based on a speciﬁc to be learned con-

cept. The elaboration questionnaire format for adding

questions includes different ﬁelds such as: question

ﬁeld, correct answer/s ﬁeld, elaboration ﬁeld, cov-

ered concepts and syllabus ﬁelds, difﬁculty level ﬁeld.

The course’s instructor analyses the proposed ques-

tion and sends individual feedback to each student. In

this way, by using the designed platform, students are

challenged not only to process the content in order to

develop speciﬁc questions for a test, but, at the same

time, they are challenged to elaborate the rationale on

which the questions and the correct answer are based

on, following in this way the elaboration principle of

effective learning.

3.2 Closer Abides to Retrieval Practice

Principle

The retrieval practice learning strategy, also known

as practice testing, involves recalling information in

low or no-stakes contexts for formative purposes. It

includes also various forms of testing that students

can engage with independently. This strategy not

only shows knowledge of the information, but also

strengthens and expands it. The efﬁcacy of testing as

a learning tool is based on numerous studies, such as

(Runquist, 1983), (Henry L. Roediger and Karpicke,

2006) and (Zaromb and Roediger, 2010). Moreover,

research states that testing a subset of information can

also impact memory for related but untested informa-

tion (Chan, 2009), (Chan, 2010), (Chan et al., 2006),

(Cranney et al., 2009), (Dunlosky et al., 2013).

The Closer platform was designed to offer a con-

text for students by self-testing the learned content.

The tests consist of multiple choice questions de-

veloped by students and analysed and approved by

teachers. The platform allows the combination of

these questions using different criteria – level of dif-

ﬁculty, the speciﬁcity of the content. In the valida-

tion phase of the platform, the students practice self-

testing, where each practice test is accompanied by

feedback involving the presentation of the correct an-

swer.

3.3 Closer Abides to Feedback Principle

The feedback strategy improve learning by reveal-

ing students about their strengths and weaknesses and

also helps them to become more aware of their own

learning process (Agarwal and Henry L. Roediger,

2018). In addition, it is strongly advised to incor-

porate feedback into practice testing as research in-

dicates that it can protect students to repeat errors

when responding to practice tests (Butler and Roedi-

ger, 2008).

Feedback is another strategy on which our pro-

posed platform was built on. Thus, the Closer plat-

form was designed so that students obtained feed-

back in different moments along their learning pro-

cess: when the multiple choice questions were ﬁnal-

ized, individualized feedback was delivered by teach-

ers in order to help students to reﬁne their questions;

at the end of each self-test practice, feedback was of-

fered regarding correct answers, the quantity of cov-

ered content, the position of student in his/her group

hierarchy based on his/her testing results.

4 COLLABORATIVE LEARNING

The term of collaborative learning can be found in

a variety of activities where students, or both teach-

ers and students involve their powers in solving prob-

lems, understanding new concepts or developing new

products (Smith and MacGregor, 1992). Collabora-

tive oriented learning environments no longer aim for

individual performance, but rather focus on fostering

a culture of teamwork, cooperation and team growth.

In this respect, collaborative learning contributes on

shaping students on two dimensions: information

gaining and social skills development (Johnson and

Johnson, 1984).

Given the beneﬁts of collaborative learning, this

section concentrates in presenting two collaboration

mechanisms: student-teacher and student-student col-

laboration, and their integration as part of the pro-

posed E-Learning software system. Although it is

described in more depth as part of Section 5, one of

the core features of the proposed platform is the ques-

tion proposal system. Through this feature, students

can contribute to their course by proposing multiple

ICSOFT 2023 - 18th International Conference on Software Technologies

722

choice questions which, once accepted, are added to

the course question data-pool.

4.1 Student-Teacher Collaboration

The ﬁrst collaborative aspect of the question proposal

system is built upon the student-teacher collaboration.

Once a student proposes a new question, the ques-

tion review process begins. Among the persons in-

volved in the review process are the course coordi-

nators (involved actors described also in Subsection

4.2). Hence the ﬁrst collaborative process is inte-

grated as part of the proposed E-Learning software

system. Through the question proposal system, be-

sides the class contribution aspect, the solution also

aims for students to develop their skills and abilities

in terms of writing on the studied topic in a profes-

sional manner. In this respect, students collaborate

with teachers in order to achieve the best version of

a question before it gets added to the course question

data-pool.

Once a student submits a question for review,

teachers can view the course pending questions on the

Question Data-Pool section from the course dedicated

page within the web platform. Regardless if a student

proposed is directly accepted (without further mod-

iﬁcations required), or not, the student will receive

feedback from teachers. However, the collaborative

mechanism is put in place when a question requires

modiﬁcations before it gets added to the course data

pool of questions (illustrated in Figure 1).

Figure 1: Teacher-student Collaboration.

4.2 Student-Student Collaboration

Up until now, the student-teacher collaborative com-

ponent of the question proposal system has been pre-

sented. In this respect, throughout this part the focus

will shift towards the second collaborative compo-

nent of the question proposal system: student-student

collaboration. In a similar fashion, certain students

can review and send feedback to questions proposed

by their classmates. This collaborative process is in-

spired by the peer review which happens in software

development teams (and not only), when a team mem-

ber will perform code review for a task implemented

by another team member. As part of designing an ef-

ﬁcient collaborative process of type student-student

one should ask himself/herself the following: Should

all students be reviewers? and What makes a student

suitable to become a reviewer?. Therefore, as part of

creating an efﬁcient learning environment, the ques-

tion proposal system of the proposed software system

is designed based on the following concepts: course

eligibility criteria and enforced reviewer mode.

The course eligibility criteria is a mechanism

through which teachers decide the minimum require-

ments for a course participant to become a reviewer.

Teachers are allowed to build their criteria according

to their liking, being able to create simple conditions

such as: ”student should have at least one accepted

proposed question”, to more complex conditions con-

taining conditionals. As for what an eligibility criteria

can be, the proposed solutions features the following

options: Points Criteria, Proposed Questions Criteria

and Compounded Criteria (using the logical opera-

tors: and, or). All of this is available in the Class

Customization page of the a coordinated course.

The idea that there is no one-size-ﬁts-all type of

system is also valid for the proposed solution, there-

fore the course eligibility criteria may not be suit-

able for all students. In this respect, the enforced re-

viewer mode is enabled for special cases. Through

this feature teachers can make some of their students

reviewers, even if they do not yet fully meet the el-

igibility criteria. Based on observation, teachers can

see whether some students already have the necessary

skills required for performing a good question review.

In this respect, teachers are allowed to enforce the re-

viewer quality for a student on a course. In order to

design a fail safe mechanism, teachers are also pro-

vided with the possibility of withdrawing the enforced

reviewer status of a student.

The student-student collaboration process can in-

crease the culture of teamwork and help students in

developing social skills and abilities that will serve

them later in life. Besides the long term beneﬁts,

through this system the overall waiting times for a

question to be added to the course data pool could

also be reduced. Nevertheless, the course data pool

is constantly updated, which means that students can

generate tests based on a wider range of questions.

Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and ChatGPT

723

5 CLOSER: AN E-LEARNING

SOFTWARE SYSTEM

As part of this paper a possible solution to be used

in higher education (but not only) as an E-Learning

software system is being presented. This section will

cover the key features of the proposed E-Learning

tool, delve into the platform’s architectural aspects,

and outline the essential elements that contribute to

an authentic learning experience.

5.1 Closer Overview

The proposed E-Learning software, Closer, is built

upon the concepts of active learning (Freeman et al.,

2014; J., 2011) and collaborative learning. The plat-

form is built based on the following ideology: each

organization (in this scenario: universities) has afﬁli-

ated teachers and students registered on the platform.

Each teacher can create courses and let other teach-

ers join their course as coordinators, while students

enroll in the taken university courses. Courses have

their syllabus attached, bringing transparency for stu-

dents and each lecture is present in the platform as a

module. Each module has a list of attached keywords

through which a student can easily understand the

main covered topics of that module. The collabora-

tive aspect of the platform has already been described

in more depth as part of Section 4 and it consists of

the multiple choice question proposal system. The

platform comes as a helping tool for students in their

preparation for exams, in this respect each student

or teacher proposed question is gathered in a ques-

tion data pool from which students can generate tests.

Each question from the data-pool has an attached list

of keywords describing the main covered topics and at

the quiz generation step, students can customize tests

to cover desired class covered topics.

5.2 Closer Architecture

The proposed E-Learning system is entirely hosted

on cloud, making use of the Amazon Web Services

(AWS) cloud provider services. However, as part

of this subsection the focus will be centered on pre-

senting the main application components, making

abstraction -as much as possible- of the cloud ser-

vices involved in building the application. Figure 2

presents a simpliﬁed view over the system’s architec-

ture, which is composed of: web client, data layer (or

the API), serverless component and the storage layer.

By far, the most important component is the data

layer, which consists of a Django based REST API

application and is the core of the proposed software

system. Through this layer, the web client is provided

with all the required information and functionalities.

The data layer assures data persistence by communi-

cating with the Postgres database and object storage

through an AWS S3 dedicated bucket. This layer also

handles the integration of the system with other sys-

tems, such as OpenAI and the serverless component.

More about the serverless component is described as

part of Subsection 5.4 and the OpenAI integration is

described as part of Subsection 5.5.

Figure 2: Closer: Simpliﬁed Architecture Diagram.

5.3 Closer Infrastructure Overview

When talking about a software system such as the

proposed one, another important aspect to be men-

tioned is the underlying infrastructure and the exist-

ing deployment processes. As part of Subsection 5.2,

it was mentioned that the proposed E-Learning soft-

ware system is cloud based, however the underlying

used cloud services were not mentioned. In this re-

spect, through Figure 3 the services and processes in-

volved in creating a production ready ecosystem for

the proposed system are emphasized.

Figure 3: Closer: Infrastructure & CI/CD Processes.

Before jumping into the particularities of some in-

ICSOFT 2023 - 18th International Conference on Software Technologies

724

frastructure components, one of the main problems

that arise in infrastructure building is the problem

of efﬁcient infrastructure management. In this re-

spect, some DevOps best practices from the industry

are integrated in the proposed software system: In-

frastructure as Code (IaC) and Continuous Integra-

tion/Continuous Development (CI/CD) pipelines. In

a more formal way, one could understand by IaC as

the process whose aim is to achieve infrastructure au-

tomation by means of techniques used in software de-

velopment (Morris, 2016). This underlying used tool

for IaC is Terraform, through which the infrastructure

corresponding to the proposed system is conﬁgured

via HashiCorp Conﬁguration Language (HCL) con-

ﬁguration ﬁles. HCL is similar to JSON and is the un-

derlying conﬁguration language used by Terraform.

By using the Terraform tool, one could easily manage

the infrastructure for a cloud based application, while

spinning up an entire environment in cloud is as done

via a simple terraform apply command.

As for the underlying CI/CD involved processes,

the proposed system present three main pipelines: In-

frastructure Deployment Pipeline, Back-End Deploy-

ment Pipeline and Front-End Deployment Pipeline.

The ﬁrst pipeline has the role of creating/updating

the underlying cloud infrastructure, while the last

pipelines build and deploy on cloud the REST API

and web client. In this respect, through CI/CD prac-

tices one could enhance the speed of delivering new

features by means of automating some steps, such

as software delivery and testing (Fitzgerald and Stol,

2017; Kumar and Mishra, 2016).

Nevertheless, some other particularities that come

across the infrastructure of the proposed software sys-

tem are the following: dockerized back-end appli-

cation and serverless component. Docker is a Plat-

form as a Service product which enhances the devel-

opment of a software product. Through virtualization

the application run inside a docker container becomes

platform independent, thus a more efﬁcient develop-

ment process is created. As for the serverless compo-

nent, the proposed solution runs a Python script inside

an AWS Lambda Function. Serverless components

take out the responsibility of infrastructure manage-

ment, while it also enhances easiness in scalability

and lower response times. Thus, the serverless com-

ponent contributes to a better overall user experience

and faster development.

5.4 An Authentic Learning Experience

Voice assistants have rapidly become part of our daily

lives, providing convenience and ease in performing

various tasks. From controlling lights to making calls,

these voice-powered devices have simpliﬁed our rou-

tines. In this respect, the focus will now shift towards

one of the most important components of the pro-

posed E-Learning software system, namely the Closer

Voice Assistant, which is powered Alexa (Amazon

powered voice assistant).

The decision of integrating a voice assistant as

part of the proposed E-Learning system was born

based on the desire of creating a learning experience

that is more relaxant and appealing for students. Tra-

ditional web applications are mostly based on visual

interaction and user manual input, which can be in-

convenient for simple actions. In this respect, by in-

tegrating the platform with the Alexa voice assistant,

students can now obtain their desired information by

means of simple voice commands.

Given its current state, the commands available

through the voice assistant can be classiﬁed into two

categories based on their scope: general class com-

mands and quiz taking commands.

The ﬁrst category of commands, the general class

commands, are commands through which the stu-

dents can quickly get information with respect to

their academic status. Such commands include list-

ing taken courses, next class on a day’s schedule, or

retrieving current study level on a class (element of

gammiﬁcation). Through such commands students

reduce unnecessary interaction with the web platform,

enhancing the overall learning experience and reduc-

ing time spent on ﬁnding desired information.

The second category of commands consists of

integrating the examination system with the Alexa

voice assistant. The quiz generation ﬂow is illustrated

through ﬁgure 4. However, students can not fully cus-

tomize Alexa generated tests, which is possible on the

dedicated web platform. In this respect, as a way of

overcoming the earlier mentioned drawback, a new

set of commands have been introduced and allow stu-

dents to take through Alexa platform generated tests.

Figure 4: Closer Voice Assistant: Quiz generation ﬂow.

Besides bringing an authentic learning experience,

the quiz taking via Alexa can help students in getting

more comfortable with the oral exams, which were

more common during the pandemic times.

Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and ChatGPT

725

5.5 Acknowledging the Power of AI

Regardless of ones personal preferences, artiﬁcial in-

telligence is experiencing a rapid surge in popularity

and utilization across various domains. In this re-

spect, it is important to acknowledge the beneﬁts that

AI usage can have in education. Thus, the question

proposal system is integrated with the engine that also

powers the newest miracle of AI: ChatGPT, an Ope-

nAI product launched in late November 2022.

The integration of the proposed E-Learning soft-

ware system with OpenAI aims to enhance the stu-

dent skills in question proposal, helping them to un-

derstand how to classify a question based on a de-

ﬁned difﬁculty scale, understand the covered topics

of a question and many more. This feature enables

the concept of personalized tutoring (Baidoo-Anu and

Owusu Ansah, 2023) with instant feedback, which al-

most impossible in asynchronous activities such as

the one modeled through the question proposal sys-

tem. It is important to allow students to explore what

the world has to offer, rather than restricting them in

a rigid system.

On a technical level, ChatGPT uses a Genera-

tive Pre-training Transformer model, which is a deep

learning neural network usually used for natural lan-

guage processing and it proves to be efﬁcient for text

generation, language translations and many others

(Haleem et al., 2022). However, no AI product is per-

fect, and ChatGPT is no exception. Among some of

the acknowledged problems of ChatGPT one can spot

biased responses (due to training data set containing

biased data) and bad responses on simple questions.

Thus, usage of such a tool requires increased atten-

tion from the user.

Given the beneﬁts that such an integration brings

to the proposed E-Learning software system, it is time

to shift the focus towards what this integration actu-

ally consists of. The proposal of a new question con-

sists of 3 steps: selecting the target modules covered

by the question, building the question statement and

answer building step. In this respect, the last two steps

were the ones that have been integrated with OpenAI.

As part of the question creation step, students have

to complete the question statement, assign a level of

difﬁculty for that question and select the main cov-

ered topics based on a provided list of keywords. If

the student did not overuse the OpenAI integration,

a message stating the possibility of using that feature

will be shown. If enabled, the system will automati-

cally generate live suggestions regarding the suitable

level of difﬁculty for that question and the covered

topics (based on the available topics). Furthermore,

if the proposed question is vague and the system can

not clearly state a level of difﬁculty, the system will

automatically suggest a list of possible answers for

which the proposed question has the suggested difﬁ-

culty level. In a similar fashion, the OpenAI integra-

tion can be used during the answer creation step.

6 CONCLUSIONS AND FUTURE

WORK

The rapid increase in technology development comes

with speciﬁc environments for educational purposes.

Events such as Covid-19 forced the education system

to change the way of teaching. These constraints cre-

ate gaps between how teachers transfer knowledge to

students and how students acquire that knowledge.

To counter these limitations, both student-centered

teaching methods and educational platforms that in-

tegrate these methods are needed.

The proposed solution takes into consideration the

three main principles that lead to efﬁcient learning

and models active and collaborative learning through

the question proposal system and the quiz taking sys-

tem. Moreover, the authenticity of the platform and

the overall learning experience is given by integrating

the Alexa voice assistant and empowering students to

use, in a controlled environment, some of the func-

tionalities provided by ChatGPT, an OpenAI product.

As for the future work with respect to this solu-

tion, one important step will be marked by validating

the proposed solution based on the ongoing research

study by analysing data gathered from the handed

surveys. Moreover, integrating the platform through

multiple university classes, would be another target

with respect to the future of this project.

Lastly, further developing the proposed learning

model is also an aspect to be considered for future

work. In this respect, one possible addition would be

enabling students to collaboratively build questions

within the platform.

REFERENCES

Agarwal, P. K. and Henry L. Roediger, I. (2018). Lessons

for learning: How cognitive psychology informs

classroom practice. Phi Delta Kappan, 100(4):8–12.

Alameri, J., Masadeh, R., Hamadallah, E., Ismail, H. B.,

and Fakhouri, H. N. (2020). Students’ perceptions

of e-learning platforms (moodle, microsoft teams and

zoom platforms) in the university of jordan education

and its relation to self-study and academic achieve-

ment during covid-19 pandemic. Journal ISSN,

2692:2800.

ICSOFT 2023 - 18th International Conference on Software Technologies

726

Baidoo-Anu, D. and Owusu Ansah, L. (2023). Education in

the era of generative artiﬁcial intelligence (ai): Under-

standing the potential beneﬁts of chatgpt in promoting

teaching and learning. Available at SSRN 4337484.

Bishop, C. F., Caston, M. I., and King, C. A. (2014).

Learner-centered environments: Creating effective

strategies based on student attitudes and faculty re-

ﬂection. Journal of the Scholarship of Teaching and

Learning, 14(3):46–63.

Butler, A. and Roediger, H. (2008). Feedback enhances

the positive effects and reduces the negative effects

of multiple-choice testing. Memory & Cognition,

36:604–616.

Chan, J. C. (2009). When does retrieval induce forgetting

and when does it induce facilitation? implications for

retrieval inhibition, testing effect, and text processing.

Journal of Memory and Language, 61(2):153 – 170.

Chan, J. C. K. (2010). Long-term effects of testing on the

recall of nontested materials. Memory, 18(1):49–57.

PMID: 19953422.

Chan, J. C. K., McDermott, K. B., and Roediger, H. L. I.

(2006). Retrieval-induced facilitation: Initially non-

tested material can beneﬁt from prior testing of related

material. Journal of Experimental Psychology: Gen-

eral, 135(4):553–571.

Clinton, V., Alibali, M. W., and Nathan, M. J. (2016).

Learning about posterior probability: Do diagrams

and elaborative interrogation help? Journal of Ex-

perimental Education, 84(3):579–599.

Cranney, J., Ahn, M., McKinnon, R., Morris, S., and Watts,

K. (2009). The testing effect, collaborative learn-

ing, and retrieval-induced facilitation in a classroom

setting. European Journal of Cognitive Psychology,

21(6):919–940.

Devi, K. S. and Aparna, M. (2020). Moodle–an effective

learning management system for 21st century learn-

ers. Alochana Chakra Journal, 9(6):4474–4485.

Doyle, T. (2008). Helping students learn in a learner-

centered environment: A guide to facilitating learning

in higher education. Stylus Publishing, LLC.

Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J.,

and Willingham, D. T. (2013). Improving students’

learning with effective learning techniques: Promising

directions from cognitive and educational psychology.

Psychological Science in the Public Interest, 14(1):4–

58. PMID: 26173288.

Fern

andez, A. R., Gonz

alez, F. S., Merino, P. J. M., and

Kloos, C. D. (2017). A data collection experience with

canvas lms as a learning platform. In LASI-SPAIN,

pages 109–123.

Fitzgerald, B. and Stol, K.-J. (2017). Continuous software

engineering: A roadmap and agenda. Journal of Sys-

tems and Software, 123:176–189.

Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K.,

Okoroafor, N., Jordt, H., and Wenderoth, M. P. (2014).

Active learning increases student performance in sci-

ence, engineering, and mathematics. pages 50 – 62.

Greving, S., Lenhard, W., and Richter, T. (2020). Adaptive

retrieval practice with multiple-choice questions in the

university classroom. Journal of computer assisted

Learning, 36(6):799–809.

Haleem, A., Javaid, M., and Singh, R. P. (2022). An era

of chatgpt as a signiﬁcant futuristic support tool: A

study on features, abilities, and challenges. Bench-

Council transactions on benchmarks, standards and

evaluations, 2(4):100089.

Henry L. Roediger, I. and Karpicke, J. D. (2006). The power

of testing memory: Basic research and implications

for educational practice. Perspectives on Psychologi-

cal Science, 1(3):181–210. PMID: 26151629.

J., R. (2011). An experimental study of differences in

study habits and long-term retention rates between

take-home and in-class examinations. International

Journal of University Teaching and Faculty Develop-

ment, 2(2):121–129.

Johnson, D. W. and Johnson, R. T. (1984). Cooperation in

the classroom, edina minnesota. A publication Book

Company.

Kumar, D. and Mishra, K. K. (2016). The impacts of test au-

tomation on software’s cost, quality and time to mar-

ket. Procedia Computer Science, 79:8–15.

Mays, B. R., Yeh, H.-C., and Chen, N.-S. (2020). The ef-

fects of using audience response systems incorporat-

ing student-generated questions on eﬂ students’ read-

ing comprehension. The Asia-Paciﬁc Education Re-

searcher, 29:553–566.

Morris, K. (2016). Infrastructure as code: managing

servers in the cloud. ” O’Reilly Media, Inc.”.

Rich, J. J., Colon, A., Mines, D., and Jivers, K. (2014). Cre-

ating learner-centered assessment strategies for pro-

moting greater student retention and class participa-

tion. Frontiers in Psychology, 5(1):1–3.

Runquist, W. (1983). Some effects of remembering on for-

getting. Memory & Cognition, 11:641–650.

Serban, C. and Vescan, A. (2019). Advances in design-

ing a student-centered learning process using cutting-

edge methods, tools, and artiﬁcial intelligence: an e-

learning platform. In Proceedings of the 1st ACM SIG-

SOFT International Workshop on Education through

Advanced Software Engineering and Artiﬁcial Intelli-

gence, pages 39–45.

Smith, B. L. and MacGregor, J. T. (1992). What is collabo-

rative learning.

Weinstein, Y., Madan, C., and Sumeracki, M. A. (2018a).

Teaching the science of learning. Cogn. Research,

3(2):1–17.

Weinstein, Y., Madan, C. R., and Sumeracki, M. A. (2018b).

Teaching the science of learning. Cognitive research:

principles and implications, 3:1–17.

Zaromb, F. and Roediger, H. (2010). The testing effect in

free recall is associated with enhanced organizational

processes. Mem Cogn, 38:995–1008.

Closer: A Tool Support for Efﬁcient Learning Integrating Alexa and ChatGPT

727