Activities and Trends Analytics in a Widget based PLE using Semantic

Technologies

Selver Softic

, Behnam Taraghi

and Laurens De Vocht

Department for Social Learning, Graz University of Technology, Graz, Austria

iMinds, Ghent University, Gent, Belgium

Keywords:

PLE, Semantic Web, Learning Analytics.

Abstract:

We report about work in progress on tracking the activities and trends of users from logs in a widget based

Personal Learning Environment (PLE) using semantic technologies and standards for retrieval. As input for

the observations, we are using the data from our self developed PLE with around 4000 active users. Last

two years we logged their activities and modeled them with RDF (Resource Description Framework)

∗

as base

for improvement analysis of existing system. The main objective of this work is to outline how learning

environments like PLE can beneﬁt from Semantic Web and its contribution for such efforts like analytics,

proﬁling, recommendations and usability.

1 INTRODUCTION

Emergence of the Web 2.0 (O’Reilly, 2005) intro-

duced participation of users as part of the web. The

transformation of internet from consuming into inter-

action medium goes hand in hand with the advances

of the web technologies. These changes also inﬂu-

ences how we think, inform ourselves, organize our

everyday activities but also how we learn. Several re-

search studies have been carried out to analyze how

Web 2.0 applications such as Blogs (Farmer, 2005;

Holzinger et al., 2009; Ebner et al., 2007), Wikis (Au-

gar et al., 2004), Podcasting (Towned, 2005) as well

as Microblogs and Social Networks generally(Ebner

and Maurer, 2009; Ebner et al., 2010a) inﬂuence

users and can enhance education. Various studies on

Web 2.0 usage amongst students (Ebner and Nagler,

2010) outline how hard it is to follow the trends and

even more to monitor them in an appropriate way.

Mashups (Tuchinda et al., 2008) and personalization

can be used to manage this challenge in learning en-

vironments. Nowadays, with increasing number of

smart phones the online presence is getting more in-

tensive for a huge population of users. They share dif-

ferent resources and contribute to the Web with their

mobile devices. This trend applies also for teachers

and learners in context of E-Learning (Ebner et al.,

2008). Those activities consider also Web 2.0 appli-

∗

http://www.w3.org/RDF/

cations and services raised like YouTube (for shar-

ing Videos), Flickr (for sharing pictures), Slideshare

(for sharing presentations), Scribd (for sharing docu-

ments), Delicious (for sharing bookmarks) etc. The

huge amount of such applications and their usage

in learning and teaching has changed the online be-

havior and attitude of learners in respect to the new

arising technologies (Downes, 2005). This led to

the idea of Personal Learning Environment (PLE),

where tiny applications (widgets) can be integrated

and combined within a learning environment man-

aged by the learners according to their actual personal

needs. Such approach resembles to the mobile ap-

plication environments in many ways, i.e. a widget

store is offered where the learners can install widgets

on one or many spaces or personal desktops. Due

to the fact that mobile technologies and social web

are available ubiquitously as well pervasively used,

they have inﬂuence our every day life and learning

environments (Holzinger et al., 2005; Klamma et al.,

2007). It is quite challenging for education not to

be overwhelmed by all these various opportunities

within a learning environment. Todays learning pro-

cess became more individual, multi facetted and ac-

tivity driven with the tendency to ad-hoc initiated col-

laboration and information exchange. All these pa-

rameters increase the complexity of online learning

platform design and organization. Dynamics involved

in this process require nowadays shorter optimization

cycles in adaptation process of Learning Management

199

Softic S., Taraghi B. and De Vocht L..

Activities and Trends Analytics in a Widget based PLE using Semantic Technologies.

DOI: 10.5220/0004414701990203

In Proceedings of the 5th International Conference on Computer Supported Education (CSEDU-2013), pages 199-203

ISBN: 978-989-8565-53-2

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: PLE Desktop at Graz University of Technology.

Systems and Personal Learning Environments. In or-

der to provide the learners an attractive surrounding

and to tackle the named problems use of learning an-

alytics regarding activities and trends for optimization

of learning process and design of learning surround-

ing emerges as the time passes by. Such data con-

tributes to the personalization and adaptation of the

learning process and belonging hosting environments.

For these purposes we logged in anonymous way user

data regarding activities on widgets in our PLE for

last two years and modeled them into learning context

using RDF in order to generate statistics that would

help us to improve the concept of the widget based

PLE. For querying the the context we used semantic

retrieval standard SPARQL

In following sections we will offer a short overview

over related work and explain how we managed to

model the activities in order to track them. We will

also show some preliminary results on current pro-

duction system. This paper will be concluded by the

discussion about the ﬁrst results and some future work

announcements.

2 RELATED WORK

The main idea of using and developing a widget based

PLE at Graz University of Technology

was to com-

bine and integrate existing university services (Ebner

and Taraghi, 2010) as well as resources and services

on the World Wide Web in one platform and in a per-

sonalized way (Ebner and Taraghi, 2010). It bases

on meshup of widgets (Taraghi et al., 2009a; Taraghi

et al., 2009b; Taraghi et al., 2009c) that represent

the resources and services integrated from the World

Wide Web within the PLE. On the other hand Web

provides lots of different services; each can be used

as supplement for teaching and learning. The PLE

has been redesigned in 2011, using metaphors such

as apps and spaces for a better learner-centered appli-

cation and higher attractiveness (Ebner et al., 2010b;

http://www.w3.org/TR/rdf-sparql-query/

http://ple.tugraz.at

Taraghi et al., 2012). A sample of PLE Desktop with

Widgets can bee seen in ﬁgure 1. The PLE has been

running since two years. In order to enhance PLE in

general and improve the usability as well as useful-

ness of each individual widget a tracking module was

implemented (Taraghi et al., 2011). Different works

outlined the importance of tracking activity data in

Learning Management Systems (Santos et al., 2012;

Verbert et al., 2011). None of them addressed the is-

sue of intelligently structuring learner data in context

and processing it to provide a ﬂexible interface that

ensures maximum beneﬁt from collected information.

The Semantic Web standards like RDF and SPARQL

where data is structured and queried as graphs and

projected on speciﬁc knowledge domain using ade-

quate ontologies has been fairly successful used to

generate correct interpretation of web tables (Mulwad

et al., 2010) to advance the learning process (Prinsloo

et al., 2012; Jeremi

c et al., 2012) as well to support

the controlled knowledge generation in E-learning en-

vironments(Softic et al., 2009). The retrieval standard

provided by Semantic Web named SPARQL enables

easily querying of semantically enriched data. This

potential was also recognised by current research in

the EU project Intelligent Learning Extended Organ-

isation (IntellLEO)

which produced in the published

ontology framework: ActivitiesOntology

to model

learning activities and events related to them along

with the surrounding environment and Learning Con-

text Ontology

which offers formalization of learning

context as general learning situation. Due to their ac-

curacy to the problem that is addressed by this work

these ontologies have been used to model the context

of analytic data collected from user logs in this work.

Our method is based on a tracking model as a

knowledge domain related context using ontologies

and query languages like SPARQL similar to cur-

rent research in the area of Self-regulated Learn-

ers(SRL)(Jeremi

c et al., 2012). Exploratory graph-

ics show that the sum of (web) user data on the ac-

cess paths and the linkage of the resources within

an environment(site) at a particular time window

gives sufﬁcient insight at what constitutes relevance;

important properties and linkages between data re-

sources(Siadaty et al., 2011). The overall goal of

is summarization of visualizations and evaluation of

statistic data that enable the PLE optimization and

present the research community used generic tech-

niques and metrics for problems in design and adap-

tation of learning environments.

http://intelleo.eu

http://www.intelleo.eu/ontologies/activities/spec/

http://www.intelleo.eu/ontologies/learning-context/spec/

CSEDU2013-5thInternationalConferenceonComputerSupportedEducation

200

3 ANALYTICS FROM LEARNERS

LOGS

3.1 Modelling User Activities

The main objective for tracking is appropriate mod-

elling since RDF offers only the framework how the

data is aligned and organized in such constructions.

This task concerns mostly the choice of the right

vocabulary or ontology. A bunch of experience is

usually necessary to complete such effort especially

when ontology has to be designed on your own. In

current research in IntellLEO EU project however

this objective has been practically implemented. One

of the main goals of this project according to the

statement from project page is building an innovative

ontological framework for learning representation

which includes learners, context and collaboration

models, serving to achieve the targeted synergy

In the realm of the IntellLEO project inside the

provided ontology framework two special ontologies

are eminent for current work. The ﬁrst is the Activity

Ontology which offers a vocabulary to represent

different activities and events related to them inside

of a learning environment with possibility to describe

and reference the environment (in this case PLE)

where these activities occur. The second contribution

from current Ontology research work in IntellLEO

project is the Learning Context Ontology which

describes the context of a learning situation. We

used as it will be shown in following sections this

Ontology to reﬂect the user activity context.

Formulation in listing 1 depicts an instance of

lc:LearningContext class in compact N3 RDF No-

tation derived from our tracking module that stores

them into a RDF Store and makes them accessible

via relying SPARQL Endpoint. Translated into nat-

ural language this instance from listing 1 reﬂects that

a ao:Logging event which tracked the learning ac-

tivity of ao:Viewing by certain anonymous um:User

inside the learning widget named LatexFormulaToP-

ngWidget as ao:Enviroment at certain time point. As

shown in this sample modeling example vocabularies

and ontologies which ﬁt appropriate for the special

case can enable a high level of expressiveness in a

very compact manner.

3.2 Querying of User Activities Trends

In order to retrieve the data relevant for the analytics

RDF instances that reside in a RDF Store at our PLE

http://intelleo.eu/index.php?id=5

Listing 1: Sample model of a learning context in N3 nota-

tion.

@preﬁx ao: <http://intelleo.eu/ontologies/activities/ns/> .

@preﬁx foaf: <http://xmlns.com/foaf/0.1/> .

@preﬁx lc: <http://www.intelleo.eu/ontologies/learning−context/ns/> .

@preﬁx rdfs: <http://www.w3.org/2000/01/rdf−schema#> .

@preﬁx um: <http://intelleo.eu/ontologies/user−model/ns/> .

<https://ple.tugraz.at/ns/activity/#Viewing>

a ao:Viewing .

<https://ple.tugraz.at/ns/users/#FSKSN>

a um:User;

foaf:name ”FSKSN” .

<http://ple.tugraz.at/ns/events/log/#7912>

a ao:Logging;

ao:performedBy <https://ple.tugraz.at/ns/users/#FSKSN>;

ao:timestamp ”2012−10−04T07:52:52” .

<https://ple.tugraz.at/ns/widgets/#LatexFormulaToPngWidget>

a ao:Enivironment;

rdfs:label ”LaTeXFormulaPNG Converter” .

<http://ple.tugraz.at/ns/learningcontext/#7912>

a lc:LearningContext;

lc:activityRef <https://ple.tugraz.at/ns/activity/#Viewing>;

lc:environmentRef

<https://ple.tugraz.at/ns/widgets/#LatexFormulaToPngWidget>;

lc:eventRef <http://ple.tugraz.at/ns/events/log/#7912>;

lc:userRef <https://ple.tugraz.at/ns/users/#FSKSN> .

environment SPARQL query language has been used.

Operability over the data is much easier then in the

case if the log data would be stored in speciﬁc struc-

ture without standardization. In this way we are able

to answer the questions like ”Show me the a monthly

activity intensity for year 2012?”. Listing 2 represents

exactly the question stated in the manner of SPARQL

syntax. The advantage of this approach is that data

formulations are ﬂexible and tolerant against any ex-

tension of representation schema, which means that

adding supplementary properties would not change

the expressiveness and retrieval of already existent in-

formation. Further since Semantic Web support the

Open World Assumption (OWA), an answer whether

or not is something reproducible out of the knowledge

base is guaranteed.

4 PRELIMINARY RESULTS

As preliminary result we are able to track the ac-

tivity trends overall time periods like presented in

ﬁgure 2. This violin graph depicts the visual an-

swer of the query from listing 2. We can see

that for year 2012 top three favored activities were

”Reading”,”Search” and ”Authoring” while activi-

ties like ”Quizzing”,”Computing”and ”Listening” are

least frequent ones. Also the intensity shows that as

expected that most activity happens at the beginning

and at the end of academic terms when PLE is pre-

sented in introductory lectures to the newcomers and

freshmen.

ActivitiesandTrendsAnalyticsinaWidgetbasedPLEusingSemanticTechnologies

201

Listing 2: Querying the intensity of all activities in PLE

after certain date.

PREFIX ao: <http://intelleo.eu/ontologies/activities/ns/> .

PREFIX foaf: <http://xmlns.com/foaf/0.1/> .

PREFIX lc: <http://www.intelleo.eu/ontologies/learning−context/ns/> .

PREFIX rdfs: <http://www.w3.org/2000/01/rdf−schema#> .

PREFIX um: <http://intelleo.eu/ontologies/user−model/ns/> .

SELECT ?actname, COUNT(?actname)

WHERE

{

?x rdf:type

lc:LearningContext;

lc:activityRef ?a;

lc:eventRef ?e;

lc:userRef ?u.

?e rdf:type

ao:Logging;

ao:timestamp ?date.

?a rdf:label ?actname;

FILTER ( ?date > ”2012−01−01T00:00:00Z”ˆˆxsd:dateTime )

}

Figure 2: Tracking the intensity of activities for year 2012.

We have chosen this example to demonstrate the

usefulness of semantic approach in a very simple case

where a vast of e.g. usability and improvement input

can be generated with a simple query and a visual-

ization interface. Demonstration like this reveals in a

very efﬁcient manner which potentials are hidden in

appliance of Semantic Web for learning analytics.

5 CONCLUSIONS AND FUTURE

WORK

The overview over distribution of activities can re-

ﬂect the overall interest of the learners within PLE.

It can be concluded that in case of our PLE users

are more consumers that contributers. Visualisation

of statistics can help to improve the PLE usability

in general. Activities such as e.g. ”Quizzing” and

”Listening” (from some learning object widgets) are

not quite popular. Corresponding widgets that sup-

port those activities must be revised regarding usabil-

ity. The statistics visualisation help us to gain deep

insight into the behaviour of a single user in a certain

period of time. In this simple case we demonstrated

that using semantic technologies enables the extensi-

bility of learning analytics. Our approach generates

uniform interfaces for information exchange, enables

ﬂexibility for visual evaluation, and also includes the

scalability regarding the enrichment of learning ana-

lytics data, since it is tolerable because of the RDF

to the schema changes. Our future efforts are aiming

the improvement of semantic structure data layer in

order to reﬂect as many aspects as possible. We also

want to review our widget store regarding the gen-

erated results in order to decide which widgets will

further provided in our PLE and which of them need

to be re-factored.

ACKNOWLEDGEMENTS

The research activities that have been described in this

paper were funded by Graz University of Technology,

Ghent University, iMinds (an independent research

institute founded by the Flemish government to stim-

ulate ICT innovation), the Institute for the Promotion

of Innovation by Science and Technology in Flan-

ders (IWT), the Fund for Scientiﬁc Research-Flanders

(FWO-Flanders), and the European Union.

REFERENCES

Augar, N., Raitman, R., and Zhou, W. (2004). Teaching and

learning online with wikis. In Atkinson, R., McBeath,

C., and Jonas-Dwyer, D., editors, Beyond the comfort

zone: Proceedings of the 21st ASCILITE Conference,

pages 95–104, Perth.

Downes, S. (2005). E-learning 2.0. eLearn, 2005(10).

Ebner, M., Holzinger, A., and Maurer, H. (2007). Web

2.0 Technology: Future Interfaces for Technology En-

hanced Learning? In Stephanidis, C., editor, Uni-

versal Access in Human-Computer Interaction. Appli-

cations and Services, volume 4556 of Lecture Notes

CSEDU2013-5thInternationalConferenceonComputerSupportedEducation

202

in Computer Science, chapter 62, pages 559–568.

Springer Berlin / Heidelberg, Berlin, Heidelberg.

Ebner, M., Lienhardt, C., Rohs, M., and Meyer, I. (2010a).

Microblogs in higher education - a chance to facili-

tate informal and process-oriented learning? Comput.

Educ., 55(1):92–100.

Ebner, M. and Maurer, H. A. (2009). Can weblogs and mi-

croblogs change traditional scientiﬁc writing? Future

Internet, 1(1):47–58.

Ebner, M. and Nagler, W. (2010). Has web2.0 reached

the educated top? In Proceedings of World Con-

ference on Educational Multimedia, Hypermedia and

Telecommunications 2010, pages 4001–4010, Chesa-

peake. Association for the Advancement of Comput-

ing in Education (AACE).

Ebner, M., Scerbakov, N., Stickel, C., and Maurer, H.

(2008). Mobile information access in higher educa-

tion. In World Conference on E-Learning in Corpo-

rate, Government, Healthcare, and Higher Education,

pages 777–782.

Ebner, M., Scerbakov, N., Taraghi, B., Nagler, W., and

Kamrat, I. (2010b). Teaching and learning in higher

education - an integral approach. Society for Informa-

tion Technology and Teacher Education International

Conference, pages 428–436.

Ebner, M. and Taraghi, B. (2010). Personal learning envi-

ronment for higher education-a ﬁrst prototype. Pro-

ceedings of World Conference on Educational Multi-

media, Hypermedia and Telecommunications, pages

1158–1166.

Farmer, J. (2005). Blogs @ Anywhere: High ﬁdelity on-

line communication. http://blogsavvy.net/how-you-

should-use-blogs-in-education.

Holzinger, A., Kickmeier-Rust, M. D., and Ebner, M.

(2009). Interactive technology for enhancing dis-

tributed learning: a study on weblogs. In BCS HCI,

pages 309–312.

Holzinger, A., Nischelwitzer, A. K., and Meisenberger, M.

(2005). Lifelong-learning support by m-learning: ex-

ample scenarios. eLearn Magazine, 2005(11):2.

Jeremi

c, Z., Jovanovi

c, J., and Ga

sevi

c, D. (2012). Personal

learning environments on the social semantic web. Se-

mantic Web.

Klamma, R., Chatti, M. A., Duval, E., Hans, H., Hvannberg,

T. E., Kravcik, M., Law, E., Naeve, A., and Scott, P.

(2007). Social software for life-long learning. Educa-

tional Technology and Society, 10(3):72–83.

Mulwad, V., Finin, T., Syed, Z., and Joshi, A. (2010). Us-

ing linked data to interpret tables. In Proceedings of

the the First International Workshop on Consuming

Linked Data.

O’Reilly, T. (2005). What is web 2.0 - o’reilly media.

http://oreilly.com/web2/archive/what-is-web-20.html.

Prinsloo, P., Slade, S., and Galpin, F. (2012). Learning an-

alytics: challenges, paradoxes and opportunities for

mega open distance learning institutions. In Proceed-

ings of the 2nd International Conference on Learning

Analytics and Knowledge, LAK ’12, pages 130–133,

New York, NY, USA. ACM.

Santos, J. L., Govaerts, S., Verbert, K., and Duval, E.

(2012). Goal-oriented visualizations of activity track-

ing: a case study with engineering students. In

Proceedings of the 2nd International Conference on

Learning Analytics and Knowledge, LAK ’12, pages

143–152, New York, NY, USA. ACM.

Siadaty, M., Jovanovic, J., Pata, K., Holocher-Ertl, T., Ga-

sevic, D., and Milikic, N. (2011). A semantic web-

enabled tool for self-regulated learning in the work-

place. In ICALT, pages 66–70. IEEE Computer Soci-

ety.

Softic, S., Taraghi, B., and Halb, W. (2009). Weaving so-

cial e-learning platforms into the web of linked data.

Proceedings of I-SEMANTICS, pages 559–567.

Taraghi, B., Ebner, M., and Clemens, K. (2012). Personal

learning environment – generation 2.0. In World Con-

ference on Educational Multimedia, Hypermedia and

Telecommunications, pages 1828–1835. AACE.

Taraghi, B., Ebner, M., and Schaffert, S. (2009a). Per-

sonal learning environments for higher education:

A mashup based widget concept. Proceedings of

the Second International Workshop on Mashup Per-

sonal Learning Environments (MUPPLE 2009), Nice,

France.

Taraghi, B., Ebner, M., Till, G., and M

uhlburger, H.

(2009b). Personal learning environment–a conceptual

study. International Conference on Interactive Com-

puter Aided Learning, pages 25–30.

Taraghi, B., M

uhlburger, H., Ebner, M., and Nagler, W.

(2009c). Will personal learning environments become

ubiquitous through the use of widgets? Proceedings

of I-KNOW, 9:329–335.

Taraghi, B., Stickel, C., and Ebner, M. (2011). Survival of

the ﬁttest – utilization of natural selection mechanisms

for improving ple. Proceedings of the ﬁrst Workshop

on Exploring the Fitness and Evolvability of Personal

Learning Environments, pages 4–9.

Towned, N. (2005). Podcasting in higher education,. Media

Onlinefocus 22,.

Tuchinda, R., Szekely, P. A., and Knoblock, C. A. (2008).

Building mashups by example. In Bradshaw, J. M.,

Lieberman, H., and Staab, S., editors, IUI, pages 139–

148. ACM.

Verbert, K., Drachsler, H., Manouselis, N., Wolpers, M.,

Vuorikari, R., and Duval, E. (2011). Dataset-driven re-

search for improving recommender systems for learn-

ing. In Proceedings of the 1st International Confer-

ence on Learning Analytics and Knowledge, LAK ’11,

pages 44–53, New York, NY, USA. ACM.

ActivitiesandTrendsAnalyticsinaWidgetbasedPLEusingSemanticTechnologies

203