=Paper=
{{Paper
|id=None
|storemode=property
|title=From Silos to Mashups Conformance through Widget-based Mashup Approaches to Learning Environments
|pdfUrl=https://ceur-ws.org/Vol-991/paper5.pdf
|volume=Vol-991
|dblpUrl=https://dblp.org/rec/conf/ecis/Paulsson13
}}
==From Silos to Mashups Conformance through Widget-based Mashup Approaches to Learning Environments==
https://ceur-ws.org/Vol-991/paper5.pdf
FROM SILOS TO MASHUPS
CONFORMANCE THROUGH WIDGET-BASED MASHUP APPROACHES TO
LEARNING ENVIRONMENTS
Fredrik Paulsson, Umeå University, Umeå, Sweden, fredrik.paulsson@umu.se
Abstract
Different types of digital learning environments have evolved over the years and there is an ongoing
evolvement of different concepts and technology. During the last 10-15 years Learning Management
Systems have dominated. In recent years there has been a development of concepts like Personal
Learning Environments using mashup technology. This paper presents a study of a mashup approach
to the digital learning environment using web widgets. A prototype was developed and tested, covering
some technological aspects such as modularity, integration and adaptability as well as pedagogical
aspects, such as pedagogical flexibility, technological responsiveness and pedagogues roles and
attitudes. The study shows that widgets and mashup technology can be used to construct digital
learning environments that have the potential to replace traditional LMSes and that pedagogues are
positive to the adaptive learning environments and flexibility that they represent.
Keywords: LMS, MUPPLE, web widgets, mashup, VLE, PLE, e-learning
1 Introduction
Different types of Digital Learning Environments (DLE) have evolved over the years and there is a
strong momentum and an ongoing progression of different ideas and concepts for the DLE. During the
last 15 years a lot has revolved around the idea of Learning Platforms, such as Learning Management
Systems (LMS). These systems are commonly set forth as a common solution for a range of
educational needs – much like a “Business System” for learning and education. However, LMSes have
been criticized for being inflexible and hard to adapt to different pedagogical contexts and needs (see,
e.g., (Paulsson, 2008), (Paulsson & Naeve, 2006b) and (Atwell, 2007)). LMSes are also commonly
criticized for being too much about the administrative aspects of learning and education and only little
about pedagogical activities and pedagogical processes. Hence, having a strong focus on Learning
Management rather than on the actual learning activities and pedagogical processes per se - as the term
also suggests. From a system perspective LMSes are often criticized for being designed and
implemented in a silo-like fashion, contributing to lock-in effects of information and processes,
preserving old-fashioned system borders. The criticism is actually similar to the critique often heard
about business systems in general. There is also a built-in conflict between the development and
implementation of systems like LMSes on the one hand and the development and implementation of
social software and Web 2.0 on the other hand. While many LMSes currently in use try to create a
well-defined kind of “shielded community” for learning, web 2.0 is associated with open communities,
global social interaction and open information services that can be used as building blocks for new
services1 - such as for a Personal Learning Environment (PLE), as described in (Atwell, 2007), (Jones,
2008) and (Wilson et al., 2006).
1 Observe that the notion of services for Web 2.0 refer to services that target users and are not equivalent to services as in
Service Oriented Architectures (SOA), which is a software design paradigm (Erl, 2007). While the technology platform
46
In order for such web services to be used as building blocks in mashups (i.e. a composition of
services) the building blocks need to be well defined and with well-defined interfaces. Many web 2.0
services use proprietary interfaces such as the Twitter API, the Facebook API or APIs from Google
and/or they use lightweight interfaces and protocols, such as RSS or Atom. This works well in many
cases, but in order to build more sophisticated mashups there is a need for more standardized and
sophisticated interfaces and concepts for interaction that can be generalized as exemplified in
(Paulsson & Berglund, 2008), (Paulsson, 2008), (Wilson, Blinco, & Rehak, 2004) and (Mulligan,
2009). This can obviously be accomplished by using advanced proprietary APIs, as illustrated in
(Paulsson & Berglund, 2006), but from a wider perspective, common open standards are preferable,
and in the long run a necessity for mashed-up learning environments to be practical. This is also one of
the issues the study discussed in this paper was set out to examine, addressing the question: “What are
the possibilities for constructing mashed up learning environments based on available open
standards?” The study also briefly evaluates and the relationship between mashup approaches and
adaptability/responsiveness to pedagogical requirements.
The next section briefly describes the state of art, followed by a discussion on some central concepts
and ideas, followed by a presentation of the study and the prototype implementation of a mashup
learning environment. Finally the results of the study are discussed in the light of the ongoing progress
and previous research in the field.
1.1.1 LMSes and PLEs
Simply put, a Personal Learning Environment (PLE) can be described as a learning environment
where the learner is in focus as well as in control of the learning environment. However, the main
objective of the PLE is to put the learner in control of his own learning rather than in control of the
learning environment, even though these two are obviously related. Learning is regarded as a constant,
ongoing process, as is the evolvement and change of the learning environment. The learning
environment needs to be responsive and adapted to different contexts, needs and pedagogical
requirements. There is also a focus on social interaction and the use of Web 2.o technologies in
general. These are qualities that are commonly emphasized, such as in (Paulsson, 2008), (Atwell,
2007), (Wilson et al., 2006) and )(Jones, 2008
While LMS-like system are typically implemented by most educational institutions, the movement
within the teaching community as well in the research community is towards more adaptive and
responsive learning environments, similar to PLEs, see e.g., (Atwell, 2007) (Wilson et al., 2006)
(Jones, 2008) (Severance, Hanss, & Hardin, 2010) and (Palmér, Sire, Bogdanov, Gillet, & Wild,
2009). However, while pedagogical concepts like responsive learning environments are attractive, the
technology currently in use doesn’t support it very well. At the same time, education need specialized
services for dealing with pedagogical requirements, such as Personal Development Plans (PDP),
digital portfolio, services for discovery and integration of digital learning resources, and so forth,
which are resulting in several good and useful tools for learning, but they are not well integrated with
the rest of the VLE (Paulsson, 2008) (Sultan, 2010) (Severance et al., 2010). These and similar issues
are often addressed through different approaches to system integration on a per system basis, such as
by using proprietary APIs or more general integration by Web Service technology, such as described
in (Paulsson, 2008) (Erl, 2007) (Paulsson, 2006) (Ogrinz, 2009) and (Gonzalez, Penalvo, Guerrero, &
Forment, 2009). However, such approaches to building the learning infrastructure has turned out to be
problematic for several reasons. Firstly, it becomes expensive to integrate “per system”, using
proprietary APIs. API integration tends to make the systems hard coupled, which supresses flexibility,
as described in (Paulsson, 2008) and (Brereton & Budgen, 2000). Secondly, using (commonly SOAP-
underlying Web 2.0 services may very well be a SOA platform, there is an unfortunate mix-up of those two rather different
notions of services when discussing Web 2.0.
47
based) Web Service technology tend to become very complex as well as expensive, adding an cost, as
well as technical, overhead (Paulsson, 2008) (Brereton & Budgen, 2000) and (Preciado, Comai, &
Sánchez-Figueroa, 2005) and not really suitable for the kind of integration needed for adaptive and
responsive learning environments controlled by the user. And thirdly, by mixing a monolithic concept,
like the LMS with a modular service based approach some of the technical flexibility needed for
dealing with some of the pedagogical requirements is lost, as illustrated in (Paulsson, 2008) (Atwell,
2007) (Wilson et al., 2006) and (Palmér et al., 2009).
1.1.2 Widgets and Mashups
In recent years there has been a general development on the Internet towards modularity and an
alternative kind of loosely couple services driven by less complex and more web friendly service
integration, such as using RESTFul APIs [21] and lightweight APIs and protocols, such RSS and
Atom combined with widget and mashup technologies, see e.g. (Yu, Benatallah, Casati, & Daniel,
2008) (Wong & Hong, 2008) and (Hoyer & Fischer, 2008).
In conjunction with this development, a number of standards and specification, dealing especially with
issues related to widgets and mashups, have been developed. There are several (but similar) definitions
of a mashup. A web-mashup is commonly defined as being a combination of different services on the
web in a way that create a new composite application (or service) with an added value. A widget-
based mashup obviously uses widget technology and is typically constructed using a mashup
environment, such as Netwibes or iGoogle, see e.g. (Casquero, Portillo, Ovelar, Romo, & Benito,
2008), (Godwin-Jones, 2009) and (Hoyer & Fischer, 2008). A mashup can also be created by very
simple means, using simple web tools that allow users to combine services on the web by matching
and mixing information using lightweight interfaces such as RSS or Atom. However, in such cases it
is mainly about mashing up information and not that much about mashing up functionality and
services in a way that goes beyond the delivery and consumption of information – even though
information mashups can be valuable in many cases as part of a digital learning environment
irrespective of type.
Even so, if you are a developer or an experienced user you might want to use one of the more
sophisticated approaches that are available for the development of web-based applications, or RIA
(Rich Internet Applications) (see (Preciado et al., 2005) and (Fraternali, Rossi, & Sánchez-Figueroa,
2010)), as it is sometimes referred to, that has the potential of moving computing from the desktop to
the web.
Standards are a necessity to make Widget-Mashups work in a wider perspective and in symbiosis with
the development Widget and Mashup technology, a number of standards and specifications has
evolved and most distinguished among those are most likely the Open Social specifications, described
in (Mitchell-Wong, Kowalczyk, Roshelova, Joy, & Tsai, 2007) and (Häsel, 2011), and the package
from W3C, described in (Wilson, Sharples, & Griffiths, 2008) and (Wilson, Daniel, Jugel, & Soi,
2012).
The Widget and Mashup technology concept stands out as exceedingly suitable for the next generation
of learning environments, fulfilling the flexibility requirements for personal and responsive learning
environments by providing standardized frameworks for modularity and loose integration on the web.
As discussed in (Paulsson, 2008), (Wheeler, 2009), (Verpoorten, Westera, & Specht, 2011) and
(Severance et al., 2010), there is a relation between the flexibility provided by modularity and the
pedagogical flexibility needed to make the learning environment adaptive and responsive. Hence,
using modular approaches for the design and implementation of digital learning environments can
address some of the LMS related issues that were described in the previous section.
The terms Widgets and Mashup will from now on referring to W3C or OpenSocial widgets and
Widget Mashups in the context of this paper.
48
The study described in this paper was carried out in a project called WiLearn. Before describing the
WiLearn project some brief background will be given in the next section.
1.1.3 Background and context
The WiLearn project is the successor of two previous research projects dealing with modular digital
learning environments and the relation between modularity, technical flexibility and pedagogical
adaptability and responsiveness. The first project, called the Virtual Workspace Environment (VWE)
was started in 1998. The objective of VWE was to implement a modular digital learning environment
that also moved the computer desktop to the web. The underlying idea was that teachers and students
should be able to construct and change their own, as well as their shared, learning environment by
combining building blocks (components) called VWE Tools. The learning environment didn’t exist
until someone assembled it for a specific pedagogical purpose, using components suitable for
supporting that specific pedagogical scenario. VWE was constructed using Java RMI on the server and
Java Applets on the client (in the browser) and an important part of the VWE was a collection of
server-sider services that kept the environment together by dealing with the internal communication of
components, session management and other system stuff. The technical solution and architecture made
VWE somewhat proprietary and there were no standard components that could be used out of the box.
Everything had to be adapted for VWE. The use of Java RMI and applets created an overhead that
made it to scale and it put a lot of demand on the browser, see (Paulsson & Berglund, 2003). Some of
those problems were addressed by the use of SOAP web service, but even though this solved some of
the problems, it also contributed to new overhead, see (Paulsson & Berglund, 2008).
Taken altogether this made it very hard to make comprehensive field studies with teachers and
students. However, in spite of this VWE contributed with some valuable knowledge about the distinct
relationship of technical flexibility, provided by modularity and the pedagogical flexibility and
adaptability as a result. This was discussed in detail in (Paulsson, 2008). The VWE project also
pointed out the importance of standards. General standards as well as Learning Technology Standards
(LTSC) that are needed to keep the learning processes and the digital learning content together,
something that are discussed in (Paulsson & Naeve, 2006a), (Salmenjoki, Paulsson, & Vainio, 2007),
(Nilsson, 2001) and (Varlamis & Apostolakis, 2006).
WiMupple was the name of the second project and the objective of WiMupple was to compare a
Widget-based Mashup solution to the Java RMI and SOAP-based approaches of the VWE. A simple
prototype was developed and compared to the properties of VWE. Some new conclusions could be
drawn about the importance of web friendliness and some of the findings from the VWE-project, such
as the importance of standards, could be verified once again. The main differences were less technical
complexity and overhead in the implementation and the ability to use standardized components (i.e.
widgets) from third-party. The use of the Wookie widget server to implement the W3C widget
standards made this possible; see (Wilson et al., 2008). The ability to use widgets developed by others
is important as it facilitates reuse and stimulates development. Even though the Wookie server was
used, some in-house development was needed in order to make the WiMupple platform usable. This
resulted in a couple of REST APIs based on Yahoo Querying Language (YQL) (2011) and some
proprietary functionality, which actually added some similar problem as VWE. However, this was
necessary to develop the same level of functionality as in VWE The WiMupple project was discussed
in (Paulsson, 2012).
The next part of this paper presents the WiLearn project the study carried out, followed by a
discussion of the results and some ideas for future studies.
49
2 Objective
The objective of the WiLearn project was to study how teachers approach modular digital learning
environments and to study if and how the technical flexibility offered contributes to the view of the
pedagogical possibilities. Another objective of the study was to implement a prototype of a generic
mashup environment for constructing digital learning environments based on widget and mashup
technology, without the limitations of the VWE and the WiMupple environments. The purpose was to
use the prototype for the study fulfilling the first objective.
It should be noted that the basic idea is still the same as with the previous studies: that teachers and
students should be in control and that the learning environment can be assembled for a specific
situation based on pedagogical needs and requirements. Different types of digital learning
environments can be constructed, such as PLEs or LMS-like learning environments, depending on the
type of widgets and supporting backend systems that are available and used.
3 Methodology
An experimental approach was used for the study in the WiLearn project. The approach was inspired
by the design-based research methodology described in (Wang & Hannafin, 2005) and in (Anderson &
Shattuck, 2012). The intension was to create a setting that was as similar to a real-life pedagogical
scenario as possible by developing a functional prototype that didn’t restrict the field part of the study
more than necessary.
3.1 Technology settings
An important premiss was to avoid unnecessary development and avoid developing everything from
scratch, which was the case in the VWE-project. There is a multitude of ongoing development and
projects addressing widgets and mashups and one of the first design decisions were to go as
mainstream as possible and existing work has been used whenever possible. The main platform used
was the Apache Rave framework. Apache Rave is described by the development team as a “…web
and social mashup engine that aggregates and serves web widgets…”2. The reason for the choice of
Apache Rave was mainly that, even though Apache Rave is in an early stage of development, it is
based on two solid Widget platforms providing stable widget frameworks. The main components of
Apache Rave are the Apache Wookie server, serving W3C widgets, and the Apache Shindig, serving
OpenSocial widgets, see (Wilson et al., 2008) and (Häsel, 2011). This also means that WiLearn has
the ability to use both W3C widgets and OpenSocial widgets for mashups, which increases the number
of available widgets. Another important reason for using Apache Rave is that it takes care of things
like session management, management of user workspaces and login/single sign-on (SSO). Especially
SSO is a central feature for mashups, as the user will actually be using several systems, even though it
may have the look and feel of one integrated system. Besides, Apache Rave can support different ways
of handling login, which makes it easier to integrate with existing user management and system
legacy. Figure 1 illustrates the components and architecture of Apache Rave.
Using Apache Rave as the basis, the features needed to implement a simple but functional and flexible
learning environment were developed. Both OpenSocial widgets and W3C widgets were used and the
choice of widget framework mainly depended on two things: whether there were existing widgets that
could be used and the type and character of the functionality to be implemented. It turned out that a
large part of the needed widgets were already available as widgets developed by others. However,
these often needed some adaption in order to be suitable for our purposes and to be suitable for use in
2 The Apache Rave website. Retrieved from http://rave.apache.org/ 2013-03-01
50
a learning environment. Some of the core (administrative) features that are normally found in the LMS
were implemented using widgets, but the focus of the widget implementation was on widgets adding
pedagogical value. Besides the possibility to use widgets that ere adapted by for the sake of the study,
a possibility to search two different “Widget stores” was implemented. In total over 300 000 widgets
that could potentially be found and used to add new and/or alternative functionality to the learning
environment.
The technology requirements at the client-side are very low as a consequence of going mainstream.
There is no need for software installations or plug-ins. The only demand on the client is a reasonably
new browser and a stable Internet connection. This is a distinct difference from both the VWE project
and the WiMupple project where certain browser versions and software was a requirement.
Figure 1. Components and architecture of Apache Rave3
3 The Apache Rave website. Retrieved from http://rave.apache.org/ 2013-03-01
51
3.2 The field study
The prototype environment that was described in the previous section was used for a field study. The
field study involved in total 34 pedagogues from public schools and was carried out as a number of
workshops involving 6-12 pedagogues each time. Informal conversations took place during the
workshops and from the perspective of the study, these were threated as a form of contextual
interviews, even though the form was very informal and the main activity during the workshops was
the work with the digital learning environment. The focus of the workshops was on pedagogues
composing their own digital learning environment using the WiLearn platform and widgets.
In order to get the most out of the “collective intelligence” of the pedagogues participating in the field
study, the pedagogical scenario and the composition of the learning environment was carried out as a
collective effort, with support from a developer in order to be able to encounter for issues that may
occur from the fact that the WiMupple platform is still in a prototype stage and that problems may
occur from time to time.
4 Results and discussion
First of all, it is obvious that a lot has happened technology wise since the first attempt towards a
modular digital learning environment that was made in the VWE project. The fact that there is now a
common concept a framework for modularity on the web is probably one of the most important single
factors that may change the way the construction of digital learning environments are viewed and
constructed. In the first attempt made in the VWE project, there was no other way than making the
environment proprietary. Even though it was possible to address the problem of information silos (in
part) and to work with standards to a certain level, there was still not possible to build a digital
learning environment that was completely open and based on standards. In the second attempt, in the
WiLearn project, it was possible to work with the common concept of widgets-based mashups and a
much more extensive use of standards both general technology standards, standards connected to the
widget and mashup field and learning technology standards. Still there was a need to develop some
quite extensive parts that could not be solely based on standards and common practise since the widget
and mashup frameworks were not developed enough. From that perspective the situation in the
WiLearn was a lot better than in the VWE project, but still far from ideal and there was still a need for
quite a lot of tweaking and solutions developed by the project team.
From the experiences of the WiLearn project it can be concluded that the concept of widgets and
mashups has matured and so has the related standards. The development of frameworks, such as the
Wookie and Shindig frameworks have come far and are now fully usable and almost production
stable. Even though the Apache Rave platform is still in early development it is reasonably stable.
Much due to the fact that both Apache Wookie and Shindig that are used as the basis for Apache Rave
are stable products.
From a technology point of view, the study worked well and there were only a few technical problems
and some of them could be countered for in the field study as it was carried out in a controlled
environment with members of the project team helping out. The problems that arose where mainly due
to the prototype nature of the WiLearn platform, with limited and/or missing functionality or related to
widgets developed for different versions of the standards. Some widgets found in the external widget
stores were simply not able to run on the WiLearn platform. This was an annoying problem, but far
from being a showstopper. Another limitation, related to the versioning problem, was the limited
support for inter-widget communication. Inter-widget communication is essential for creating the same
level of conceived integration as in a traditional LMS-system and for implementing advanced and rich
functionality. In a mashup everything is loosely coupled and different widgets are normally
completely self-standing and self-contained and not “aware” of the context in which they are used.
52
This makes it harder to maintain the feeling of a well-integrated learning environment. Still, the tight
integration of LMS-functionality is often seen as a core problem of the LMS and is also what much of
the criticism is targeting. With well-developed mechanisms for inter-widget communication
integration can be managed by loose coupling instead, favouring adaptability and openness. There are
strategies to deal with this problem and as the widget standards evolve, together with the
implementation of html-feature, this problem is slowly being dealt with, se e.g. (Isaksson & Palmer,
2010) and (Sire, Paquier, Vagner, & Bogaerts, 2009) for an in-depth discussion on the inter-widget
communication issue.
The WiLearn study, together with the experiences from previous studies, show that it is quite possible
to implement all of the core functionality in a modular fashion using widgets and mashup technology.
This basically means that mashup environments could potentially replace traditional LMSes. There are
no restrictions regarding the type of digital learning environment that can be composed using a
mashups. It is possible to create LMS-like learning environments as well as personal learning
environments of combinations of the two. Everything depends on the widgets that are used in the
mashup and the structure that is implemented. Hence, with the right set of widgets, a complete LMS
could be built using as a mashup, even though an LMS may not be what is wanted or needed. The
technical development is fast and by choosing modularity and standards in favour of monoliths, it is
likely to be much more feasible to keep up with new and emerging technology. In fact, even though
educational institutions in general are not among the early adopters of technology, many teachers are
and everyone gains from an infrastructure for learning that is designed for change and that can meet
with early adopters.
The study also confirms that modular learning environments are better adapted to suit different
learning theories and pedagogical approaches as well as to different and changing pedagogical
scenarios. Such features are beneficial and even essential in many learning scenarios, especially when
working with pedagogical methods and approaches like Problem Based Learning (PLE) or case
methodology, where it is hard (if not impossible) to foresee the learning path from start to finish
beforehand – and thereby also to foresee the needs of the learning environment. Adaptability is also an
essential property as it allows for the learning environment to be distributed (service-wise) over the
Internet. At the same time it makes it possible to personalize and adapt the learning environment at the
service level for group preferences as well as for personal preferences. The correlation between
modular environments and adaptability and responsiveness, as well as technologies ability to match
different learning theories, were discussed in detail in (Paulsson, 2008).
Several pedagogues also asked for ways to change the overall structure and look and feel of the
WiLearn environment. This is an area that needs to be addressed in future versions.
Many pedagogues feel (and in fact are) restricted by the LMS and they got creative when presented to
the mashup environment. Interesting, as well as a bit surprising, was that when pedagogues were given
the freedom to influence how the learning environment was designed and what functions that should
be available there was still a focus on the administration of learning and education. The difference was
that the focus was on the small administrative tasks that is a part of the daily classroom activities and
needed for communication with pupils and parent. This kind of support is often quite simple and
surprisingly enough often missing in LMSes. Another, but related issue was the ability to create a
personal “toolbox” with personal tools needed to facilitate teaching and all that is related. A common
requirement was widgets that integrate different cloud services, such as Google Apps, DropBox,
Twitter and likewise. There are widgets available that can do this, but the quality is varying and most
of them are rather unsophisticated. The quality issue created a problem in general when searching
widgets outside the WiLearn widget store. There is a huge amount of widgets available in different
widget stores and many of them are of poor quality and they are adapted different versions of the
standards, which cause some widgets not to function properly in the WiLearn environment. There is
no good way of searching only for widgets intended for learning and education.
53
Another type of widgets that were asked for by the pedagogues participating in the study was support
for specific pedagogical tasks and needs, such as widgets that could replace software for learners with
special needs or widgets performing isolated tasks – i.e. replacing pedagogical desktop software. This
is interesting and such widgets would in many cases be reasonably straightforward to develop.
Unfortunately there are only a few widgets available of this kind and they are often dependent of
language and cultural factors. In general, the supply of widgets that are especially developed for
learning and education is small and of varying quality. Altogether, this contributes to an interesting
business case where market competition is opened up for smaller actors to compete with LMS vendors
by providing small and specialized components acting as building blocks in a mashup learning
environments. In fact, it makes the learning environment independent of a specific LMS vendor to
implement certain functionality. It has proven to be quite straightforward to develop simple widgets
that can act as clients to different legacy (as well as to other) systems. The relation of system legacy
and WiMupple is a concern that were raised a couple of times during the study and the idea of widgets
acting as clients proved to be working, at least for simple cases.
In all, the study presented in this paper has contributed to the research field by providing a better
understanding for how modularity and widget-based mashups influence the way that pedagogues
approach the use of ICT and digital learning environments.
5 Future research
There are some potentially intriguing developments around the corner that are likely to benefit the
development of mashup environments. On the one hand there is the general development, such as the
gradual evolvement of html5 and standards for widgets and mashups. On the other hand there are
developments within the field of learning technology standards, which at least on paper, look very
promising from a modularity point of view. Among the most interesting efforts are the new
specifications from IMS and especially the content standard IMS Common Cartridge (IMS CC) (2008)
and the IMS Learning Tools Interoperability (IMS LTI) (2010) specifications. We are currently in the
process of examining whether IMS CC can be used as a packaging format for WiLearn and
furthermore, if IMS LTI can be used as a standard for widget communication and interactions related
to pedagogical activities within a widget-based learning environment – as a complement to general
standards. Severance et al. describes some experiments in (Severance et al., 2010) where IMS LTI was
tested in a mashup environment and the results are promising and could be taken even further in the
context of the WiLearn project.
The intention is to develop the WiLearn environment into a “production stable” framework for
composing mashed-up digital learning environments that can be implemented and tested in a real-life
educational scenario as a substitute to the LMS and to be able to study the use and effects in a real
context over a longer period of time.
Another direction, that has already started, is the integration of the Spider and the WiLearn
environment. The Spider is a national search service for digital learning resources that connects a
number of repositories, using either metadata harvesting or federated search. This makes it possible to
search for learning resources from several sources from a single point (Paulsson, 2009). The idea is to
use the Spider as a widget store as well and by that making it possible to search for widgets and
learning content in the same context.
54
References
Atwell, G. (2007). Personal Learning Environments - the future of eLearning? Retrieved 6 June, 2009
from www.elearningeuropa.info/files/media/media11561.pdf
Brereton, P., & Budgen, D. (2000). Component-based systems: a classification of issues. Computer,
33(11). doi:10.1109/2.881695
Casquero, O., Portillo, J., Ovelar, R., Romo, J., & Benito, M. (2008). iGoogle and gadgets as a
platform for integrating institutional and external services. Proceedings from Mash-Up Personal
Learning Environments. Proc. of 1st Workshop MUPPLE.
Isaksson, E., & Palmer, M. (2010). Usability and Inter-widget Communication in PLE. Proceedings
from Mashup Personal Learning Environments 2010, MUPPLE'10, Barcelona, Spain.
Erl, T. (2007). SOA Principles of Service Design (1 ed.). Boston, MA: Prentice Hall.
Fraternali, P., Rossi, G., & Sánchez-Figueroa, F. (2010). Rich Internet Applications. Internet
Computing, 14(3), 9-12. doi:10.1109/MIC.2010.76
Godwin-Jones, R. (2009). Emerging technologies personal learning environments. Language,
Learning & Technology, 3(9).
Gonzalez, M. A. C., Penalvo, F. J. G., Guerrero, M. J. C., & Forment, M. A. (2009). Adapting LMS
Architecture to the SOA: An Architectural Approach. Proceedings from Fourth International
Conference on Internet and Web Applications and Services. ICIW '09.
Häsel, M. (2011). Opensocial: an enabler for social applications on the web. Communications of the
ACM, 54(1), 139–144. Retrieved from http://doi.acm.org/10.1145/1866739.1866765
Hoyer, V., & Fischer, M. (2008). Market Overview of Enterprise Mashup Tools
Service-Oriented Computing – ICSOC 2008. Proceedings from Lecture Notes in Computer Science.
IMS Common Cartridge Specification. Retrieved January 07, 2009 from
http://www.imsglobal.org/cc/index.html
IMS GLC Learning Tools Interoperability Basic LTI Implementation Guide. Version 1.0 Final
Specification, 25.
Jones, D. (2008). PLES: FRAMING ONE FUTURE FOR LIFELONG LEARNING, E-LEARNING
AND UNIVERSITIES. Proceedings from Lifelong Learning Conference 2008, Yeppoon, Qld, AU.
Mitchell-Wong, J., Kowalczyk, R., Roshelova, A., Joy, B., & Tsai, H. (2007). OpenSocial: From
Social Networks to Social Ecosystem. Proceedings from Digital EcoSystems and Technologies
Conference, 2007. DEST '07.
Mulligan, C. E. A. (2009). Open API standardization for the NGN platform. Communications
Magazine, IEEE, 47(5), 108-113. doi:10.1109/MCOM.2009.4939285
Sultan, N. (2010). Cloud computing for education: A new dawn? International Journal of Information
Management, 30(2), 109-116. Retrieved from
http://www.sciencedirect.com/science/article/pii/S0268401209001170
Nilsson, M. (2001). The Semantic Web: How RDF will change learning technology standards.
Ogrinz, M. (2009). Mashup Patterns Designs and Examples for the Modern Enterprise. Addison
Wesley.
Palmér, M., Sire, S., Bogdanov, E., Gillet, D., & Wild, F. (2009). Mapping Web Personal Learning
Environments. Proceedings from Mash-Up Personal Learning Environments (MUPPLE'09).
Paulsson, F. (2006). A Service Oriented Architecture-framework for modularized Virtual Learning
Environments. In A. Mendes-Vilas, A. Solano Martin, J. Mesa Gonzáles, & J. A. Mesa Gonzáles
(Eds.), Current Developments in Technology-Assisted Education (1 ed., p. 21 62). FORMATEX.
Paulsson, F. (2012). Mashing up the learning environment - Evaluating a widget-based approach to
Personal Learning Environments. Proceedings from The Seventh International Conference on
Internet and Web Applications and Services (ICIW 2012), Stuttgart.
Paulsson, F. (2008). Modularization of the Learning Architecture: Supporting Learning Theories by
Learning Technologies. Stockholm: Royal Institute for Technology (KTH). Retrieved from
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-4712
55
Paulsson, F. (2009, May 24-28). Connecting learning object repositories - strategies, technologies and
issues. Proceedings from The Fourth International Conference on Internet and Web Applications
and Services, Venice, Italy.
Paulsson, F., & Berglund, M. (2006). A Service Oriented Architecture-framework for modularized
Virtual Learning Environments. Proceedings from m-ICTE2006, Seville.
Paulsson, F., & Berglund, M. (2008). Suggesting a SOA-framework for modular virtual learning
environments: comparing two implementation approaches. International Journal of Web-Based
Learning and Teaching Technologies, 3(1), 43-57.
Paulsson, F., & Berglund, M. (2003). SOA-frameworks for modular virtual learning environments:
comparing implementation approaches. In J. Mosemann (Ed.), Novel Approaches in Web-Based
Learning Technologies: Tools for Modern Teaching. Hershey, PA: IGI Global.
Paulsson, F., & Naeve, A. (2006a, Cunningham, Paul). Establishing technical quality criteria for
Learning Objects. Proceedings from in the proceedings of eChallenges 2006, Barcelona, Spain.
Paulsson, F., & Naeve, A. (2006b). Virtual Workspace Environment (VWE): A Taxonomy and
Service Oriented Architecture Framework for Modularized Virtual Learning Environments -
Applying the Learning Object Concept to the VLE. International Journal on E-Learning, 5(1), 45-
57.
Preciado, J., Comai, S., & Sánchez-Figueroa, C. (2005). Necessity of methodologies to model Rich
Internet Applications. Proceedings from the 2005 Seventh IEEE International Symposium on Web
Site Evolution (WSE’05).
Salmenjoki, K., Paulsson, F., & Vainio, A. (2007). Social Learning - How Web is Changing the Way
We Process Information and Knowledge. Proceedings from 5th International Conference on
Emerging e-learning Technologies and Applications (ICETA), Stará Lesná, Slovak Republic.
Wilson, S., Daniel, F., Jugel, U., & Soi, S. (2012). Orchestrated User Interface Mashups Using W3C
Widgets. In A. Harth & N. Koch (Eds.), Current Trends in Web Engineering (pp. 49-61). Springer
Berlin Heidelberg.
Severance, C., Hanss, T., & Hardin, J. (2010). IMS Learning Tools Interoperability: Enabling a Mash-
up Approach to Teaching and Learning Tools. Techn., Inst., Cognition and Learning, 7, 245-262.
Sire, S., Paquier, M., Vagner, A., & Bogaerts, J. (2009). A messaging API for inter-widgets
communication. Proceedings from the 18th international conference on World wide web, Madrid,
Spain.
Anderson, T., & Shattuck, J. (2012). Design-Based Research : A Decade of Progress in Education
Research? 41: 16-25,http://edr.sagepub.com/content/41/1/16.full. Educational Researcher, 41, 16-
25.
Varlamis, I., & Apostolakis, I. (2006). The present and future of standards for e-learning technologies.
Interdisciplinary Journal of Knowledge and Learning Objects, 2, 59–76.
Verpoorten, D., Westera, W., & Specht, M. (2011). Infusing reflective practice in eLearning courses
– can widgets help? International Journal of Technology Enhanced Learning, 3(1), 93-109.
Retrieved from http://dx.doi.org/10.1504/IJTEL.2011.039066
Wang, F., & Hannafin, M. (2005). Design-based research and technology-enhanced learning
environments. Educational Technology Research and Development, 53(4), 5-23. Retrieved from
http://dx.doi.org/10.1007/BF02504682
Wheeler, S. (2009). Learning Space Mashups: Combining Web 2.0 Tools to Create Collaborative and
Reflective Learning Spaces. Future Internet, 1(1), 3–13. doi:10.3390/fi1010003
Wilson, S., Blinco, K., & Rehak, D. (2004). An e-Learning Framework A Summary. Proceedings from
Alt-I-lab 2004.
Wilson, S., Liber, O., Johnson, M., Beauvoir, P., Sharples, P., & Milligan, C. (2006). Personal
Learning Environments: Challenging the dominant design of educational systems. Proceedings
from First European Conference on Technology Enhanced Learning (ECTEL), Crete, Greece.
Wilson, S., Sharples, P., & Griffiths, D. (2008). Distributing education services to personal and
institutional systems using Widgets. Proceedings from Mashup Personal Learning Environments
(MUPPLE08), Maastricht.
56
Wong, J., & Hong, J. (2008). What do we” mashup” when we make mashups? Proceedings from the
4th international workshop on End-user software engineering.
(2011). Yahoo! Query Language. Retrieved from http://developer.yahoo.com/yql/
Yu, J., Benatallah, B., Casati, F., & Daniel, F. (2008). Understanding Mashup Development. IEEE
Internet Computing, 12(5), 44-52. Retrieved from
http://search.ebscohost.com/login.aspx?direct=true&db=eoah&AN=17067655&site=ehost-
live&scope=site
57