Personalisation of MOOCs

The State of the Art

Ayşe Saliha Sunar

, Nor Aniza Abdullah

1,2

, Su White

and Hugh C. Davis

Electronics and Computer Science, University of Southampton, University Road, Southampton, U.K.

System and Technology, University of Malaya, 50603, Kuala Lumpur, Malaysia

Keywords: Personalisation, Adaptive Online Learning, Connectivism, MOOCs.

Abstract: Researchers in the field of educational technology are paying huge attention to the widespread adoption of

Massive Open Online Courses (MOOCs) in the study of learning online. MOOCs are discussed in many

angles including pedagogy, learning sustainability, and business model. However, there are very few

discussions around MOOCs personalisation. In this paper, it is aimed to examine and analyse the literature

on personalisation of MOOCs to identify the needs, the current states and efforts to personalise learning in

MOOCs. The findings denote that the pedagogical design of MOOCs is currently insufficient due to

massive and geographically dispersed learners with diverse educational backgrounds, learning requirements

and motivations. Many believe that personalisation could address this lacking in MOOCs. Among the most

popular services being proposed or implemented in the literature are personalised learning path,

personalised assessment and feedback, personalised forum thread and recommendation service for related

learning materials or learning tasks.

1 INTRODUCTION

Massive Open Online Courses (MOOCs) is an

emerging area in technology-enhanced learning

(Jona and Naidu, 2014). Even the first MOOCs

course, Connectivism and Connective Knowledge

08 (CCK08), has attracted thousands of learners. It

should be noted here, this online course was not

announced as a “massive open online course”, the

term “massive open online course” was first

introduced in 2008 by Dave Cormier to describe

George Siemens and Stephen Downes’ CCK08

online course (McAuley et al., 2010). The first

MOOCs course was based on connectivism theory

that addresses issues about connecting people and

resources to construct knowledge. It emphasises the

importance of providing social platforms to learners

to support their interactions with the course content,

rather than just transmitting knowledge to them

(Siemens, 2005). This kind of MOOCs is later

known as cMOOCs.

In 2011, Sebastian Thrun designed a MOOCs

course on Artificial Intelligence at Stanford

University. Pedagogically, this MOOCs course was

different from the first MOOCs. It is more teacher-

centric in which learning goals and learning plans

were predefined for potential learners. This kind of

MOOCs is named as xMOOCs, and it is based on

the behaviourist learning theory (Daniel, 2012).

Even though MOOCs is relatively a new trend in

technology-enhanced learning, concerns on teaching

and learning with MOOCs are still the same with

those on online education (Hollands and Tirthali,

2014; Shaw, 2012), for instance, how can MOOCs

be pedagogically efficient to address different needs

of its learners? Research attempts to address this

issue are discussed further in Section 3. One

proposed study is to provide MOOCs

personalisation through educational data mining in

order to improve learning experience in MOOCs. In

this paper, the state of the art of personalisation in

MOOCs based on a study on the related literatures is

presented. The methodology is presented in Section

2. Analysis and findings are reported in Section 3 in

order to identify the aspects of MOOCs’s

personalisation that are commonly addressed by

researchers and those that are still not sufficiently

look into. The existing personalisation approaches

and report of the critical reviews on them are further

investigated in the sub sections of Section 3. Based

on the findings, suggestions on ways to improve the

delivery of personalised learning in MOOCs are

provided in Section 4. Section 5 concludes the study

Sunar A., Abdullah N., White S. and C. Davis H..

Personalisation of MOOCs - The State of the Art.

DOI: 10.5220/0005445200880097

In Proceedings of the 7th International Conference on Computer Supported Education (CSEDU-2015), pages 88-97

ISBN: 978-989-758-107-6

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

and presents suggestions for future work.

2 METHODOLOGY

In order to review the literature, similar

methodology used by Liyanagunawardena et al.

(2013) and Yousef et al. (2014)'s researches is

applied. The articles between 2011 and 2014 (by

November, 30) are searched by the keywords

“MOOCs personalisation” and “adaptive MOOCs”

on several academic databases, Google Scholar, The

British Journal of Educational Technology,

American Journal of Distance Education, Journal of

Online Learning and Technology, ISI Web of

Knowledge and IEEEXplore. The reason of

choosing this particular time period is that 2011 is

the year in which both xMOOCs and cMOOCs have

been discussed (Daniel, 2012) and MOOCs has

become rapidly and widely used in online learning

as reported in (Liyanagunawardena et al., 2013). Not

only peer-reviewed articles were analysed in this

paper, but also the grey literature, for example

institutional reports were also searched and

analysed.

Table 1: The result of the search by the keyword “MOOCs

personalisation”.

Year Search result Relevant

Google Scholar

2011 17 0

2012 29 1

2013 313 11

2014 427 14

The British Journal of Educational Technology

2011 0 0

2012 0 0

2013 0 0

2014 1 0

American Journal of Distance Education

2011 0 0

2012 0 0

2013 0 0

2014 4 0

Journal of Online Learning and Teaching

2011 0 0

2012 0 0

2013 0 0

2014 0 0

ISI Web of Knowledge

2011 0 0

2012 0 0

2013 4 1*

2014 0 0

Table 1: The result of the search by the keyword “MOOCs

personalisation” (cont.).

IEEEXplorer

2011 0 0

2012 0 0

2013 0 0

2014 0 0

Table 1 and 2 illustrate the number of papers that

have been retrieved, along with the number of

relevant papers to the personalisation of MOOCs

over the years based on the searched keywords

“MOOCs personalisation” and “adaptive MOOCs”,

respectively. While the year 2012 is called and

referred many times as “the year of the MOOC”,

personalisation of MOOCs has been on the rise since

2013 (The New York Times, November 2, 2012:

http://www.nytimes.com/2012/11/04/education/edlif

e/massive-open-online-courses-are-multiplying-at-a-

rapid-pace.html?pagewanted=all&_r=0).

Table 2: The result of the search by the keyword “adaptive

MOOCs”.

Year Search result Relevant

Google Scholar

2011 19 0

2012 72 1*

2013 422 18***

2014 623 17***

The British Journal of Educational Technology

2011 0 0

2012 0 0

2013 3 0

2014 1 0

American Journal of Distance Education

2011 0 0

2012 0 0

2013 0 0

2014 2 0

Journal of Online Learning and Teaching

2011 0 0

2012 0 0

2013 0 0

2014 0 0

ISI Web of Knowledge

2011 0 0

2012 0 0

2013 3 1*

2014 4 0

IEEEXplorer

2011 0 0

2012 0 0

2013 1 1*

2014 5 3**

* 1 same result with the other search.

** 2 results of them are the same with the other search.

*** 8 results of them are the same with the other search.

PersonalisationofMOOCs-TheStateoftheArt

Figure 1 clearly illustrates that the amount of

attention for personalised learning in MOOCs is

drastically increased in the last two years.

Even though, the number of search results is over

600 papers (see Table 2), relevant papers are only a

few among them (40 papers in total). Papers on

studies regarding adaptive online education systems,

and other issues related to MOOCs are also retrieved

along with papers on mass personalisation in

MOOCs with these keywords. However, the relevant

papers only indicate studies that are based on mass

personalisation.

Figure 1: The total number of papers and relevant papers

by the searches for the keywords “MOOCs

personalisation” and “adaptive MOOCs”.

This study only considers the relevant papers for

analysis. The analysis is organised according to the

purposes and scoped of the studies, and the

personalisation or adaptation techniques used.

3 DATA ANALYSIS

Once the redundant papers are eliminated from the

collection of relevant papers, it is observed that

some papers rhetorically indicate needs for

personalisation in MOOCs while some others

attempt to develop personalisation services in

MOOCs. Therefore, the relevant papers are clustered

into three categories in this study:

1. NEEDS: Represents the ‘Need for

personalisation in MOOCs’. This category of

research papers indicates the need or

opportunity for MOOCs personalisation.

They mainly report findings that lead to the

need for personalised learning in MOOCs.

However, the papers in this category do not

propose any project, framework or system for

designing or implementing personalisation in

MOOCs.

2. PROPOSALS: Represents the ‘Plan to

implement personalisation in MOOCs’. This

category of research papers expresses ideas

and proposals for personalisation projects in

MOOCs. However, the plans for the intended

personalisation systems have not yet been

implemented.

3. IMPLEMENTATIONS: Represents the

attempts for ‘Personalisation Service in

MOOCs’. This category of papers expresses

partly or fully implemented and experimented

proposals for personalisation in MOOCs.

However, majority of researches in this

category are in progression state with no

definitive outcome yet.

Figure 2 illustrates the number of papers in each

category over the years. The figure denotes that only

one paper emphasises the need for personalisation in

MOOCs in 2012 while 2013 is the year with the

highest number of papers (13) calling for

personalisation. In 2013, there are 5 descriptive

papers on proposals for personalisation in MOOCs

but only 3 papers proposed partly or fully

personalised MOOCs functions in MOOCs learning

environment. Generally, the number of papers in

categories of Proposals and Implementations

increases in 2014 after the call for adaptive MOOCs

in the previous year. The results show that there is a

rapid growing of interest towards personalised and

adaptive learning in MOOCs. Additionally, it is

predicted that there will be more implemented and

fully experimented studies in the coming years.

Figure 2: The number of papers in each category over the

years.

3.1 Needs

Fasimpaur (2013), Freeman and Hancock (2013),

Godwin-Jones (2014), and Harman and Koohang

(2013) indicate that a huge amount of human data

CSEDU2015-7thInternationalConferenceonComputerSupportedEducation

can be collected through MOOCs. The availability

of the big data in MOOCs, and tools to perform

learning analytics would make it possible for a

personalised system to predict learners’ learning

behaviours and preferences in order to deliver

personalised learning and assistance to MOOCs

learners. Shaw also (2012) points out that this pool

of human data could be used to create a human

model in intelligent tutoring system (ITS) for

MOOCs. Similarly, Yates (2013) and Knox (2014)

highlight that data mining and data analytics for

prediction could make MOOCs adaptive. Slightly on

a different note, Kay et al. (2013) predict that

educational data mining and learning analytics

should be applied for MOOCs’s social network

analysis to enable personalised learning in MOOCs.

Kalz (2014) further supports the argument by

highlighting that these techniques could make

MOOCs a more suitable technology to support

lifelong learners.

The importance of offering personalised learning

in MOOCs is further expressed by the following

researchers. For instance, Amo (2013) believes that

MOOCs should offer student-centred learning for

effective and quality education in order to meet each

individual learner’s learning expectations in

MOOCs. However, she emphasises that current

pedagogy and design of MOOCs is not enough to

improve students’ outcomes. As there are many

exciting and available pedagogies in technology

enhanced learning such as peer assistance and

assessments, social networking, and gamification,

the author suggests for the incorporation of these

pedagogies into MOOCs. This can be accomplished

through the use of learning analytics and continuous

monitoring of students’ interactions so that

automated assessment with instant feedback can be

personalised to every student to improve quality

learning in MOOCs.

McLoughlin (2013) and Knox et al. (2014) also

address the current inefficiency of learners’

feedbacks in MOOCs. They point out that MOOCs

environment is convenient for offering personalised

contents and feedbacks to learners based on their

learning goals. This is because MOOCs provides

learning flexibility and sense of independence

between learners and teachers which are important

when implementing personalisation in technology-

enhanced learning.

Additionally, Kalz and Specht (2013) point out

that the current MOOCs design does not consider

the diversity of its learners. The authors suggest that

building sub groups that share similar attitudes and

interests could be a solution. The authors further

indicate that the heterogeneity problem in MOOCs

community is akin to the problem of learning

network. The authors describe learning network as a

connection of humans, actors, agents, institutions

and learning resources organised for a learning

program/course. To deal with diversity in learning

networks, several services for learner support in

learning networks should be utilised, such as

placement support service (navigation support), a

recommender service, and knowledge matchmaking

service. By using these intelligent personalisation

techniques, different needs and interests among

diverse learners community in MOOCs can be

addressed. To further support the importance of

addressing diversity among learners, Cavanaugh

(2013) whose work focuses on MOOCs assessments

for credits for the post secondary education, states

that personalised learning pathways for learners

could help them build their capabilities to obtain

credits.

Kizilcec et al. (2013) are concerned with low

completion rate in MOOCs. Therefore, they have

conducted a study to examine patterns of learners’

engagement and disengagement with the MOOCs

course, and consecutively they have suggested for

MOOCs to offer adaptive content or assistance to

learners according to their needs. Their suggestion

is further supported by Martin et al. (2013) who

believe that learning in MOOCs can be encouraged

by providing predefined personal path and super

badges that indicate the competence level of each

individual learner.

On the other hand, Aoki (2013) and Stine (2013)

focus on business model for MOOCs. While Stine

(2013) indicates mass personalisation can have a

positive business impact to MOOCs, Aoki (2013)

points out that MOOCs is representing a new

business model. Aoki (2013) states that content

providers for lectures, assessments/accreditation and

tutorial supports will eventually be separately

established and organised. The author presumes that

the learners’ data will be shared among separate

organisations to enable personalisation in MOOCs.

Despite the apparent needs for personalised

learning in MOOCs, Kay et al. (2013) point out that

the existing MOOCs courses are not even half way

through in implementing personalisation.

Nevertheless, without personalisation, learners may

reduce their participations and eventually drop out

from a MOOCs’s course, which is one of the biggest

concerns of MOOCs (Stevanović, 2014).

Noteworthy that even though, there is nonexistence

PersonalisationofMOOCs-TheStateoftheArt

of personalisation practice on the existing MOOCs

platforms, Hollands and Tirthali (2014) point out

that MOOCs still present the term POOC

“Personalised Open Online Course” into their full

report. It is also stated that the success of MOOCs

will depend on how much the learning process is

personalised.

3.2 Proposals

The literature that is considered under this category

mainly involves project launches which are funded

for the aim of personalising online education for

masses, projects’ proposals for implementing

personalisation in the existing non-personalised

MOOCs, and conceptual research frameworks.

Most of the research works are driven by

concerns over the inefficiency of MOOCs design,

delivery, and assessments. For instance, Daradoumis

et al. (2013) and Bassi et al. (2014) voice their

concerns in several different research papers.

According to the authors, as most of MOOCs

courses are not learner-centric, and they provide

same content for all learners, the effectiveness of the

tutoring is generally poor, feedbacks are insufficient

and peer-based evaluation is usually unprofessional.

To address these deficiencies, the authors propose an

agent-based framework for MOOCs. Agents collect

data and analyse them according to several

perspectives including educational goal, pedagogical

preferences, time management and so forth. The

analysed data is used by other agents for content

customisation, tutoring feedback, system-learner

alert as well as assessing and monitoring learners’

learning progress in MOOCs. The authors indicate

that intelligent agents could also be used for

reducing fraud and cheating during online tests.

Broun et al. (2014b) propose a personalisation

component which will be integrated to the existing

EMMA platform. EMMA platform is a MOOC

platform delivering courses in different languages

from different European Universities; therefore,

learners may be overwhelmed with huge number of

courses and language choices. Through this

personalisation component, EMMA aims to provide

personalised feedback and individualised learning

paths to support learners to achieve their learning

goals.

De Maio et al. (2014) believe that learners’

engagement with the video lecture materials in

MOOCs as passive. To improve learners’

engagement with MOOCs, the authors propose a

methodology to support learners to navigate the

fragments of one or more videos lectures so that

learners could connect their goals and prior

knowledge with the key concept of the lectures. The

authors use taxonomy building for constructing a

knowledge model for the concepts of lectures. The

main idea is to enable inter-linking between different

MOOCs courses and navigate learners to related

ones. However, this part of the research has not been

conducted.

Similarly, Wilkowski et al. (2014) have

conducted an analysis on learners’ goals and their

achievements on the tested skills and activities by

executing “Mapping with Google” course in

MOOCs. Each learner was asked to complete a

questionnaire about their learning goals to join the

course and their previous experiences with the

Google map. The authors then compared learners’

learning goals with their behaviours in the course

(i.e. watched videos, completed activities), and

found out that their behaviours were very much

determined by their goal. Therefore, the authors

conclude that the course delivery could be

personalised based on learners’ goals. Their

proposed system could be adapted to learner’s

requirements in two ways. First is to ask for

learners’ goals prior to delivering personalised

learning pathway to each of them. Secondly, to have

learners select the course elements such as some

video lectures and assessments from a list for a

customised course.

Fasihuddin et al. (2014) propose an approach for

personalised learning experience in MOOCs based

on learners’ learning styles. The authors define the

kind of material that should be included in the

lecture for a particular learning style. For example,

while visual learning objects should be accessible

for visual learners, such need is not a necessity for

verbal learners. However, this is an ongoing research

and a prototype is still not yet completed.

Elkherj and Freund (2014) have developed an

adaptive hint system for the undergraduate online

course “Introduction to probability and Statistics” on

the Webwork, which is a platform for managing

homework assignments in mathematics. This course

was attended by 176 students and hints were written

by the tutor each time learners made a mistake or

failed a test. The authors express that the need for

manual labour for analysing learners’ failure and

writing helpful hints makes the system inconvenient

for MOOCs. Therefore, they propose some possible

approaches that could address this problem. The first

is for students to hints to their peers. Secondly,

create hint libraries. Finally, use machine-learning

CSEDU2015-7thInternationalConferenceonComputerSupportedEducation

techniques to map students’ mistakes with hints and

consecutively send the most relevant hint to them.

Brouns et al. (2014a) propose ECO sMOOC for

the EU-funded project called Elearning,

Communication and Open-data: Massive Mobile,

Ubiquitous and Open Learning (ECO). sMOOC

refers to being a social-based MOOCs which is

accessible from different types of social media and

mobile devices. Learning is executed devices

through content contextualisation based on learners’

interactions and participations in the course using

mobile and gamification approaches. The ECO

sMOOC environment is described as learner-centric

approach, which is adaptable to learners’ intention.

However, the project is in the very early stage, and

any real experience with it has not yet available.

Bain et al. (2013) propose AMOOC (Accessible

Open Online Course) movement to make MOOC

courses more accessible for learners with

disabilities. The paper focuses on delivering course

content in appropriate forms for disable learners.

They also mention that the system will be conducted

using Adaptive Mobile Online Learning (AMOL)

for adapting coursework to each learner’s learning

style.

Collet (2013) proposes POEM (Personalised

Open Education for the Masses) platform project for

designing personalised learning management system

(LMS) for massive learning. The author believes

that personalisation of massive education is only

possible with intelligent ICT (Information and

Computing Technology) platforms. In POEM, visual

and dynamic Knowledge Maps of domains for each

course are constructed to provide different possible

learning paths to learners. POEM will also provide

inter-tutorship and automatic assessments. Apart

from that, the system will ask learners to post new

questions or new contents to the platform.

Bansal (2013) and Birari (2014) have utilised the

concept of ITS for personalising learning

experiences with MOOCs from different

perspectives. Bansal (2013) focuses on providing

recommendations for learners to do additional

learning activities to improve their lack of

knowledge on a particular topic. In order to model

learners’ knowledge, the author uses the fuzzy

cognitive map. On the other hand, Birari (2014)

models learners’ cognitive state by Bayesian

network so that adaptive testing and adaptive

guidance can be delivered to learners.

Slightly on a different note, Blanco et al. (2013)

has identified three weaknesses in MOOCs: high

dropout rate, lack of cooperative activities among

learners, and poor continuity of learning

communities when a MOOCs course ends.

According to the authors’ definition, learning

community includes activities, resources, and similar

groups. To improve learning experiences in

MOOCs, the authors have outlined the components

of learning community that should be personalised

based on learners’ learning goals, previous

knowledge, etc. These personalisation inputs are

captured and diagnosed through initial assessments.

Similarly, Zhuhadar and Butterfield (2014) point

out that providing a singular curriculum to a diverse

MOOCs community has caused low completion

rates in MOOCs. To address this problem, the

authors propose Personalised Open Collaborative

Courses (POCCs) which tracks learners’ attitude

during the course and delivers the personalised

content based on learners’ activities and their prior-

knowledge. In order to achieve this goal, the authors

examine sub communities in MOOCs to design a

personalised social recommender system.

3.3 Implementations

Research works reported in this category provide a

more concrete evidence of approaches towards

implementing personalisation in MOOCs, such as

early stage experimental results, a system framework

or results of system performance tests. This category

considers either partly or fully implemented

personalised systems that may have performed some

kind of testing on either system performance or

student performance. Noteworthy that majority of

the systems have not yet completed their final

evaluations, and the projects are still ongoing.

An algorithm of an adaptive study planner for

MOOCs learners, targeted to novice learners in

MOOCs is presented by Alario-Hoyos et al. (2014)

and Gutiérrez-Rojas et al. (2014a). The adaptive

planner creates a personalised study schedule for

each learner based on their priority of the course,

available time slot and the course requirements.

However, this system has not yet been evaluated.

Burgos and Corbí (2014) present a rule-based

technology-enhanced learning recommendation

model in order to improve users’ performance in

MOOCs and other Open Educational Resources

(OERs). The model tracks learners’ performances

and their interactions with the lectures. It

consecutively map the related data according to the

tutor’s rules for recommendation such as minimum

number of required activity in a lecture and

minimum score on a given test. Based on the results

PersonalisationofMOOCs-TheStateoftheArt

of rules mapping, a recommendation is made. If a

learner satisfies the tutor’s rule to be successful, then

the learner gets positive comment such as “Well

done!” and gets recommendation for the subsequent

tasks. Otherwise, the system gives alert feedback to

the learner to request support from the online tutor

and peers, and locks any further activities.

Ketamo (2014) utilises ITS technologies for

providing recommendations to support learners’

cognitive progress and motivation in MOOCs. The

content that will be provided to learners is defined as

semantic network. This approach requires a learner

to complete and succeed relevant test on a learning

concept prior to recommending the next related

learning concepts. According to the preliminary

evaluation results, learners’ performances were

improved when using the recommendation service.

However, a considerable portion of learners was still

not motivated to learn, and eventually dropped the

course.

Shatnawi et al. (2014a, 2014b) propose system

architecture for providing personalised feedback to

learners in MOOCs by using text-mining technique.

Since the course creators are not able to provide

timely feedback due to massive number of learners,

the authors propose a method for providing

automatic content related feedback by using domain

ontology, machine learning, and natural language

processing. When a learner writes a post, the system

will determine its type, whether it is a question, a

comment, or a feedback, and organised it into a

suitable domain under the related topic in a

repository. If a learner posts a question, the system

will automatically search the repository and returns

semantically relevant information or personalised

feedback to the learner.

Sonwalker (2013) proposes an adaptive MOOC

that offers adapted learning contents based on

learning styles with the concern of pedagogical

effectiveness of MOOCs. The author proposes the

learning cube that illustrates organisation of learning

objects developed in text, graphics, audio, video,

animations, and simulations according to different

learning styles. In this study, learners’ learning style

is diagnosed via a diagnostic test as suggested by

Blanco et al. (2013). The performance test result is

promising.

Yang et al. (2014) propose a personalised

support on MOOCs discussion forums for helping

learners to reach the topics in which they are

interested. The authors use both collaborative and

content filtering techniques to capture the most

relevant forum threads. Their system performance

test results show that the system performance of the

proposed personalisation model is satisfactory,

however, learners’ satisfaction test has not yet

examined.

Some researchers modify existing personalised

technology-enhanced learning systems for MOOCs

courses. For example Miranda et al. (2013)’s work

aims to provide a pedagogy-based guide for items

assessment based on the ontological relations

between learning subjects in the lectures which are

defined by the course creator. According to a

learner’s assessment’s score, a personalised learning

pathway is constructed for the learner. Similarly,

Henning et al. (2014) also adapt an existing

technology-enhanced learning system into MOOCs.

The system supports learners through personalised

navigation based on their learning performances and

the association between learning subjects.

4 DISCUSSION

Result from the analysis of the needs related

literature shows that the pedagogical design of

MOOCs is insufficient, therefore, educational data

mining should be applied to provide personalised

services such as personalised learning pathways,

personalised assessments, adaptive feedbacks, and

recommender services. To address the needs for

personalisation in MOOCs, researches in category

Proposals and category Implementations have

proposed several outlines, frameworks, and projects’

proposals, as well as prototypes for implementing

personalisation and adaptation in MOOCs.

For instance, Kalz and Specht (2013) and

Kizilcec et al. (2013) from category Needs suggest

to cluster MOOCs’s learners for personalisation. The

suggestion was implemented by Blanco et al. (2013),

Fasihuddin (2014) and Sonwalker (2014) in which

they applied a diagnostic test at the beginning of the

course to understand which group (i.e. learning

style) a learner belongs to. However, this method is

based on learners’ participations in the diagnostic

test, and majority of learners are not interested in

doing tests. Realising this problem, Zhuhadar and

Butterfield (2014) have suggested using some social

networking analysis (SNA) techniques to diagnose

learners and automatically cluster them according to

the most suited sub community in MOOCs based on

their activities. Even though this method does not

need learner’s self-statement, a learner is required to

participate in the course’s lectures and activities

until the system can gather sufficient information

CSEDU2015-7thInternationalConferenceonComputerSupportedEducation

about the learner in order to determine a suitable

cluster for the learner.

Another example is by the work of Shaw (2012)

who believes that the application of ITS technique

can actualise mass personalisation in MOOCs. The

belief was translated by Bansal (2013), Bariri (2014)

and Ketamo (2014) who implemented ITS

techniques in MOOCs for personalising contents,

learning pathways, and providing recommendations.

Note that even though Yang et al. (2014) and

Brouns et al. (2014a) did consider the social feature

of MOOCs, for example they personalise online

forum threads to learners based on their forum

activities and peers connections, they did not build a

personalised learning network in MOOCs or social

network analysis for improving learning networks as

suggested by Kalz and Specht (2013) and Kay et al.

(2013). Therefore, continuity problem of learning

communities identified by Blanco et al. (2013)

remains unsolved.

5 CONCLUSION AND FUTURE

WORKS

In conclusion, this literature survey has

demonstrated that there is a growing trend of

researchers embarking in the possibility of

implementing personalisation and adaptation in

MOOCs in order to improve users’ engagements,

hence reduce MOOCs’ drop-out rate problem. The

trend is mainly motivated by the fact that MOOCs’s

learning has the potential to spark demands for

personalised learning due to its massive and

geographically dispersed learners with diverse

background. In addition to that, MOOCs

environment does provide the basic requirements for

personalised learning such as the availability of huge

learners’ data, flexible learning, and learner-teacher

independence. Our categorisation of the literature

identified three distinct types of papers. 1) These

concerned with the need or motivation for

personalisation in MOOCs. 2) Outlines of plans or

proposals for implementing personalisation in

MOOCS. 3) Accounts and evaluations of the

implementation of personalisation services in

MOOC. We found that data mining techniques are

often used to exploit huge learners’ data in MOOCs,

and majority of the studies are concerned on the

pedagogical design issues. Therefore, many

researchers have proposed solutions based on

personalisation and adaptation techniques such as

personalised learning pathways and personalised

feedback. However, there is not yet any tangible

research that focuses on building personalised

learning networks even though the need has been

identified by Kalz and Specht (2013), Kay et al.

(2013) and Blanco et al. (2013). It is expected that

this issue will gain more attention in the nearest

future.

REFERENCES

Alario-Hoyos, C., Leony, D., Estévez-Ayres, I., Pérez-

Sanagustín, M., Gutiérrez-Rojas, I., and Kloos, C. D.

(2014). Adaptive planner for facilitating the

management of tasks in MOOCs. In Proceedings of

the V Congreso Internacional sobre Calidad y

Accesibilidad de la Formación Virtual, CAFVIR 2014,

Antigua Guatemala, Guatemala, pp. 517-522.

Amo, D. (2013, November). MOOCs: experimental

approaches for quality in pedagogical and design

fundamentals. In Proceedings of the First

International Conference on Technological Ecosystem

for Enhancing Multiculturality, pp. 219-223, ACM.

Aoki, K. (2013) Paradoxes between Personalisation and

Massification: The Future of Education, Conference

Proceedings 2013. 3

Edition.

Bain, K., Chan, B., and Bates, L. (2013). AMOOC:

Improving Access to MOOCs using Speech

Recognition. Retrieved from http://

liberatedlearning.com/wp-content/uploads/2011/02/

AMOOC_Apr_2013.pdf.

Bansal, N. (2013). Adaptive recommendation system for

MOOC (Doctoral dissertation, Indian Institute of

Technology, Bombay).

Bassi, R., Daradoumis, T., Xhafa, F., Caballé, S., and

Sula, A. (2014). Software Agents in Large Scale Open

Elearning: A Critical Component for the Future of

Massive Online Courses (MOOCs). SINCOS 2014,

Salerno, Italy. September 10-12, 2014. In proceedings

of the Sixth IEEE International Conference on

Intelligent Networking and Collaborative Systems, pp.

184-188. IEEE Computer Society.

Birari, N. (2014). Intelligent Tutoring System using

Computerised Adaptive testing and interaction logs

for MOOCs (Doctoral dissertation, Indian Institute of

Technology, Bombay).

Blanco, Á. F., García-Peñalvo, F. J., and Sein-Echaluce,

M. (2013, November). A methodology proposal for

developing adaptive cMOOC. In Proceedings of the

First International Conference on Technological

Ecosystem for Enhancing Multiculturality, pp. 553-

558, ACM.

Brouns, F., Mota, J., Morgado, L., Jansen, D., Fano, S.,

Silva, A. and Teixeira, A. (2014a). A networked

learning framework for effective MOOC design: the

ECO project approach. In A. M. Teixeira, & A. Szücs

(Eds.), 8th EDEN Research Workshop. Challenges for

Research into Open & Distance Learning: Doing

PersonalisationofMOOCs-TheStateoftheArt

Things Better: Doing Better Things, pp. 161-171.

Budapest, Hungary: EDEN. Oxford, United Kingdom.

Brouns, F., Tammets, K., and Padrón-Nápoles, C. L.

(2014b). How can the EMMA approach to learning

analytics improve employability?. Retrieved from

http://dspace.learningnetworks.org/bitstream/1820/554

2/1/submission9_M4WS.pdf.

Burgos, D., and Corbí, A. (2014). A recommendation

model on personalised learning to improve the user’s

performance and interaction in MOOCs and OERs.

UNESCO Institute for Information Technologies in

Education. IITE 2014 International Conference. Oct

14th-15th, 2014. Moscow, Russia.

Cavanaugh, J. (2013). The Coming Personalization of

Postsecondary Education Competencies. CAEL 2013

Forum & News: Competency-Based Education, pp. 2-

Collet, P. (2013). POEM (Personalised Open Education

for the Masses). Retrieved from http://

blog.educpros.fr/pierredubois/files/2013/07/Financem

ent.pdf.

Daniel, J. (2012). Making sense of MOOCs: Musings in a

maze of myth, paradox and possibility. Journal of

Interactive Media in Education, 3.

Daradoumis, T., Bassi, R., Xhafa, F., and Caballé, S.

(2013, October). A review on massive elearning

(MOOC) design, delivery and assessment. In P2P,

Parallel, Grid, Cloud and Internet Computing

(3PGCIC), pp. 208-213. IEEE.

De Maio, C., Loia, V., Mangione, G. R., and Orciuoli, F.

(2014). Automatic Generation of SKOS Taxonomies

for Generating Topic-Based User Interfaces in

MOOCs. In Open Learning and Teaching in

Educational Communities, pp. 398-403, Springer

International Publishing.

Elkherj, M. and Freund, Y. (2014, March). A system for

sending the right hint at the right time. In Proceedings

of the first ACM conference on Learning@ scale

conference: pp. 219-220. ACM.

Fasihuddin, H. A., Skinner, G. D., and Athauda, R. I.

(2014) Personalizing Open Learning Environments

through the adaptation to Learning Styles. ICITA

2014, 9

International Conference on Information

Technology and Applications, Sydney, Australia, July

2014. ISBN: 978-0-9803267-6-5.

Fasimpaur, K. (2013). Massive and Open. Learning and

Leading with Technology March/April 2013, pp. 12-

17.

Freeman, M. A. R. K., and Hancock, P. H. I. L. (2013).

Milking MOOCs: Towards the right blend in

accounting education. Academic Leadership Series, 4,

86-100.

Godwin-Jones, R. (2014). Emerging Technologies Global

Reach and Local Practice: the Promise of MOOCs.

Announcements & Call for Papers, 5.

Gutiérrez-Rojas, I., Alario-Hoyos, C., Pérez-Sanagustín,

M., Leony, D., and Delgado-Kloos, C. (2014a).

Scaffolding Self-learning in MOOCs. Proceedings of

the Second MOOC European Stakeholders Summit,

EMOOCs, 43-49.

Gutiérrez-Rojas, I., Leony, D., Alario-Hoyos, C., Pérez-

Sanagustín, M., and Delgado-Kloos, C. (2014b)

Towards an Outcome-based Discovery and Filtering of

MOOCs using moocrank.

Harman, K., and Koohang, A. (2013) MOOC 2050: A

Futuristic Tour. Issues in Information Systems, 14 (2),

pp. 346-352.

Henning, P. A., Heberle, F., Streicher, A., Zielinski, A.,

Swertz, C., Bock, J., and Zander, S. (2014).

Personalized Web Learning: Merging Open

Educational Resources into Adaptive Courses for

Higher Education. Personalization Approaches in

Learning Environments, 55.

Hollands, F. M. and Tirthali, D. (2014, May). MOOCs:

Expectations and reality. Full report. Center for

Benefit-Cost Studies of Education, Teachers College

Columbia University. Retrieved from http://

cbcse.org/wordpress/wp- content/uploads/2014/05/

MOOCs_Expectations_and_Reality.pdf.

Jona, K., and Naidu, S. (2014). MOOCs: Emerging

Research. Distance Education, 35(2), pp.141-144.

Kalz, M. (2014) (in press). Lifelong Learning and its

support with new technologies. In Smelser, N. J. and

Baltes P. B. (Eds.). International Encyclopaedia of the

Social and Behavioral Sciences. Pergamon: Oxford.

Kalz, M., and Specht, M. (2013). If MOOCs are the

answer-did we ask the right questions. Implications for

the design of large-scale open online courses.

Maastricht School of Management in its series

Working Papers, 2013/25.

Kay, J., Reimann, P., Diebold, E., and Kummerfeld, B.

(2013). MOOCs: So Many Learners, So Much

Potential... IEEE Intelligent Systems, 28(3), pp. 70-77.

Ketamo, H. (2014, June). Learning Fingerprint: Adaptive

Tutoring for MOOCs. In World Conference on

Educational Multimedia, Hypermedia and

Telecommunications, Vol. 2014, No. 1, pp. 2458-

2467.

Kizilcec, R. F., Piech, C., and Schneider, E. (2013, April).

Deconstructing Disengagement: Analyzing Learner

Subpopulations In Massive Open Online Courses. In

Proceedings of the third international conference on

learning analytics and knowledge: pp. 170-179. ACM.

Knox, J. (2014). From MOOCs to Learning Analytics:

Scratching the surface of the 'visual'. eLearn,

2014(11), 3.

Knox, J., Ross, J., Sinclair, C., Macleod, H., and Bayne, S.

(2014). MOOC Feedback: Pleasing All the People?.

Invasion of the MOOCs, 98.

Liyanagunawardena, T. R., Adams, A. A., and Williams,

S. A. (2013) MOOCs: A systematic study of the

published literature 2008-2012. The International

Review of Research in Open and Distance Learning,

14(3), pp. 202-227.

Martín, S., Peire, J., and Castro, M. (2013) Proyecto

WePrendo. Retrieved from http://www.ieec.uned.es/

CSEDU2015-7thInternationalConferenceonComputerSupportedEducation

Investigacion/archivos/informe%20WEPRENDO_DI

EEC.pdf.

McAuley, A., Stewart, B., Siemens, G., & Cormier, D.

(2010). The MOOC model for digital practice.

University of Prince Edward Island, 33.

McLoughlin, C. E. (2013, June). The pedagogy of

personalised learning: exemplars, MOOCS and related

learning theories. In World Conference on Educational

Multimedia, Hypermedia and Telecommunications,

Vol. 2013, No. 1, pp. 266-270.

Miranda, S., Mangione, G. R., Orciuoli, F., Gaeta, M., and

Loia, V. (2013, October). Automatic generation of

assessment objects and Remedial Works for MOOCs.

In Information Technology Based Higher Education

and Training (ITHET), pp. 1-8. IEEE.

Shatnawi, S., Gaber, M. M., and Cocea, M. (2014a). Text

stream mining for Massive Open Online Courses:

review and perspectives. Systems Science & Control

Engineering: An Open Access Journal, 2(1), pp. 664-

676.

Shatnawi, S., Gaber, M. M., and Cocea, M. (2014b).

Automatic content related feedback for MOOCs based

on course domain ontology. In Intelligent Data

Engineering and Automated Learning–IDEAL 2014:

pp. 27-35. Springer International Publishing.

Shaw, C. (2012). Intelligent Tutors and Personalized

Education. Retrieved from http://isites.harvard.edu/

fs/docs/icb.topic1283433.files/Shaw_12_rs.pdf.

Siemens, G. (2005). Connectivism: A learning theory for

the digital age. International journal of instructional

technology and distance learning, 2(1), pp. 3-10.

Sonwalkar, N. (2013, September). The First Adaptive

MOOC: A Case Study on Pedagogy Framework and

Scalable Cloud Architecture—Part I. In MOOCs

Forum, Vol. 1, No. P, pp. 22-29.

Stevanović, N. (2014) Effects Of Motivation On

Performance Of Students In MOOC. SINTEZA 2014

Internet and Education. DOI:10.15308: pp. 418-422.

Stine, J. K. (2013, April). MOOCs and executive

education. In Presented at the Directors Conference.

Retrieved from http://uniconexed.org/2013/research/

UNICON-Stine-Research-06-2013-final.pdf.

Wilkowski, J., Deutsch, A., and Russell, D. M. (2014,

March). Student skill and goal achievement in the

mapping with google MOOC. In Proceedings of the

first ACM conference on Learning@ scale conference,

pp. 3-10. ACM.

Yates, R. (2013). Educational Technologies to Support

New Directions in Teaching Practice. International

Journal of Information & Education Technology, 3(6).

Yang, D., Piergallini, M., Howley, I., and Rose, C. (2014).

Forum thread recommendation for massive open

online courses. In Proceedings of 7th International

Conference on Educational Data Mining.

Yousef, A. M. F., Chatti, M. A., Schroeder, U., Wosnitza,

M., and Jakobs, H. (2014): MOOCs—a review of the

state-of-the-art. Proceedings of the CSEDU 2014

conference, 3, pp. 9-20.

Zhuhadar, L., and Butterfield, J. (2014). Analyzing

Students Logs in Open Online Courses Using SNA

Techniques. The 20th Americas Conference on

Information Systems, AMCIS 2014, Savannah,

Georgia, USA, August 7-9, 2014.

PersonalisationofMOOCs-TheStateoftheArt