Learning Activity

by Chris Kew, Sheila Macneill, Lisa Corley, Oleg Liber, Paul Hollins

1. Introduction

Throughout the brief but busy history of e-learning, the learning object has occupied a central role in the conception, design and production of electronically delivered courses. As the world of e-learning evolves, the primacy of the learning object is increasingly brought into question by advocates of a view which favours activity over objects. At the heart of this activity-centred view is the assertion that while the learning object has been a staple of e-learning since its inception, it has systematically failed to actively engage the learner in anything beyond electronic page-turning and the passive consumption (but not necessarily digestion) of knowledge [Koper and Tattersall 2005]. The activity-centred view on the other hand, places the performance of individual and group activities at the fore of the e-learning experience and makes use of the learning object as an auxiliary, intended to facilitate the various forms of interaction (negotiation of meaning, etc.) at various points throughout the learning activity.

A move towards standardizing this activity-based approach to learning exists in the form of the IMS Learning Design Specification, The development of the IMS Learning Design Specification (IMS LD) [1] was a direct response to a recognised need for a specification which would allow an interoperable way of representing and sharing pedagogically complex models of learning activities, which could include collaborative activities. The Open University of the Netherlands (OUNL) [2] had been working towards creating a language to enable the sharing of learning activities and had developed the Educational Modelling Language (EML) [3] . This was used as the basis for the development of the IMS specification. The OUNL and the UK based CETIS (Centre for Educational Technology Interoperability Standards) [4] were the primary architects and driving forces behind the creation of the IMS specification.

IMS LD does not define or recommend any particular pedagogical model, instead it provides a high level language, or meta-model, that can describe many different models and how people perform activities using resources (including materials and services), and how these three things are coordinated into a learning flow [Currier and Jeffrey 2005]

While the uptake of IMS LD is still in its infancy, it has caused wide spread interest from the eLearning community due to its pedagogical underpinning. Initiatives such as the UNFOLD (Understanding New Frameworks of Learning Design) project [5] , recent European Commission 6^thFramework project, UNFOLD, a partnership between the Universitat Pompeu Fabra, OUNL, CETIS and EUCEN have helped to raise awareness of the specification. More generally, UNFOLD provided a European platform for the wider context of designing for learning, an umbrella term for any form of pedagogical planning that espouses the activity-centric point of view for which adherence to the specification is not strictly necessary. Designing for learning is part and parcel of everyday teaching practice. IMS LD provides a significant step towards rendering the design process more transparent to practitioners. This can help to facilitate the planning and execution of teaching and learning, in an e-learning context, with a shift in emphasis away from content delivery to one that promotes a greater understanding of the importance of learning activities.

2. State of the art concerning standards and interoperability

As the number of e-learning standards and specifications proliferate, the importance of the role of standardization in the field is growing. For an increasing number of designers, open standards are a sine qua non due to the degree of interoperability that they bring to e-learning. However, open standards notwithstanding, claims that interoperability is achieved at the cost of pedagogical ingenuity [Koper and Olivier 2004] have prompted the development of the IMS LD specification, an offshoot of the Open University of the Netherlands's Educational Modelling Language (EML). While space precludes an in-depth analysis of the workings of the IMS LD, a brief overview will serve to highlight the significance of this specification in the context of activity based learning.

Since its initial publication in 2003, IMS LD has distinguished itself by virtue of its high level language which, in theory, allows for the description of just about any conceivable pedagogical model for use with both individuals and groups. IMS LD is a language intended for use in modelling units of study. In effect the language affords designers the opportunity to describe the interaction and coordination of three key components in the learning process as conceived in LD, namely; the learner who interacts with resources during the course of a given activity. Through the use of the metaphor of a theatrical play, learning scenarios are described in terms of plays, acts and role-parts in such a way that both learning services and content can be organized into a user specified sequence before they are assigned to learners. Another noteworthy characteristic of IMS LD includes its various levels: A, B and C, a design feature intended to facilitate the implementation process. Level A lies at the heart of Learning Design and includes the means with which to add and coordinate people, activities and resources. Level B allows for more complex designs through the use of properties and conditions. The former are used to collect learner data including preferences, test results etc. while the latter serve to constrain the learning flow in accordance with predetermined criteria designed to accommodate a specific set of circumstances or preferences. Level C adds ever increasing sophistication to learning designs through the use of notifications which serve to automatically trigger new activities upon the completion of tasks.

From an interoperability perspective, the interest in IMS LD lies in its ability to afford learners the opportunity to work on the same activities using the same services and resources across a variety of IMS LD compliant learning platforms. Compliancy can be achieved by any system through the adoption of an IMS LD import and export format and is not dependent on the use of any particular infrastructure or methodology. The advantages of such a high level of interoperability include the ability for designers and teachers alike to export learning activities created in one VLE for use in another. At the time of writing, work is being undertaken to develop IMS LD import and export capability in a number of VLEs including Moodle, .LRN and LAMS.

Despite the innovative nature of IMS LD, the specification itself is designed in such a way that it is able to support and make use of a number of existing specifications necessary to model interoperable learning activities and assessments. These include IMS Content Packaging (IMS CP) [6] , IMS Question and Testing Interoperability (IMS QTI) [7] , IMS Simple Sequencing (IMS SS) [8] to name but a few. Indeed, in the case of the Simple Sequencing, similarities with Learning Design are such that some clarification is required to help distinguish the one from the other. The main source of confusion here resides in the fact that both specifications make use of sequencing mechanisms, but whereas Simple Sequencing is geared towards modelling single learner situations, IMS LD is based on a multi learner model (although it also has the ability to model scenarios for individuals). Moreover, it should be noted that Learning Design is endowed with more sophisticated sequencing mechanisms which allow conditions and actions to be set by the designer during design time unlike those in Simple Sequencing which are predefined. No comparative overview of Learning Design would be complete without mention of ADL's SCORM (Shareable Content Object Reference Model) [9] and SCORM 2004 which also has a number of points in common with IMS LD including the use of Content Packaging and Learning Object Metadata (LOM)[10].. But in spite of these surface similarities, their respective differences underline their variance in purpose. Such differences include SCORM 2004's limited description of runtime behavior for the single user only which, while resulting in limited pedagogical use, makes less demands of the server than IMS LD. On the other hand, SCORM profiles a large set of specifications which render it highly specific and, for this reason, easier to implement than IMS LD. In terms of expressivity however, SCORM is severely restricted in contrast to IMS LD, which allows for the extraction of pedagogy from content such that it is able to offer the kind of functionality that characterizes structured activity-based learning (e.g. peer collaboration, formative assessment, tracking and sequencing of individual and group work etc.). There has been some discussion on the possible integration of SCORM and IMS LD [Kraan 2003] although, at the time of writing, no concrete action has been taken in this regard.

While IMS LD has been shown to have the potential to bring the learning activity to the fore of e-learning, there has been some concern at the relative lack of standards related activity since the first release of IMS LD in 2003. Recent developments with IMS QTI 2.0 however look set to expand IMS LD (See Section 4)

IMS LD was intended to put pedagogy before technology by providing ways to create, run and share any number and kind of complex learning activities. It was also an attempt to address the gap between the, instructional design led specifications and profiles such as IMS Simple Sequencing and SCORM and the more constructivist approaches desired particularly from the education sector. This approach brought about a considerable level of interest from educational practitioners. However at the time of the release of the specification there was little, if anything in the way of tools and concrete examples of compliant learning designs.

This lack of implementation can probably be attributed to two main factors: firstly the specification was arguably ahead of its time - the formal orchestration of people, activities and services was not commonplace within an elearning context; secondly, the lack of tools to initially create learning designs and then run them in a web-based teaching and learning environment. Any tools that were available required a high level of XML programming capability, which was not (and still is not) a skill held by mainstream educators or indeed many learning technologists.

One system that did emerge fairly soon after the release of the IMS LD specification was LAMS (Learning Activity Management System) [11] . Developed at Macquarie University in Australia, LAMS provided a 'learning design inspired' authoring and runtime environment. The simple, drag and drop user interface provided an easily understood, teacher centric, introduction to the general principles of designing collaborative learning sequences. However, despite its apparent user friendliness, the LAMS system did not provide an interoperable solution as it did not adhere to the IMS specification. Nevertheless, it did provide a route for the less technically focused to join the wider design for learning debate. (It should also be noted that LAMS can now export LAMS sequences as IMS Learning Design, Level A packages.)

At the time of writing there is still relatively little in the way of implementation of IMS LD compliant tools, but that is changing. Key stakeholders are providing commitment to IMS LD and providing funding for further developments including the development of authoring and runtime tools which will be discussed in more detail in section 6.

We are now approaching a period where there are a number of IMS LD tools available and this coupled with projects funded by for example the Joint Information Systems Committee (JISC ) [12] in the UK should provide a basis for developing more implementation practice (see section 4).

Comments from Steve. (1) There is a danger of implying that new pedagogical thinking is confined to constructivism. Maybe we need to recognise the relevance ofcommunities of practice and argue that the shift required is from Object to Activity toInteraction. (2)Is it worth commenting on the discourse gap between those constructingLearning Designtools and approaches and those who design learning?

3. Requirements/use cases

Requirements and use cases are an invaluable means of communicating stakeholder needs to systems developers for future design considerations and modification. They can be also be used to help specify other aspects of interaction taking place within a learning activity.

The need for tools for use by both designers and teachers has been at the forefront of IMS LD based discussion since the launch of the specification. In order to make IMS LD as accessible as possible, it was apparent that any tools needed to be user-friendly, and not require an in-depth knowledge of either XML or the specification. Early IMS LD compliant tools have required a high degree of both. It became clear that some intermediary phase was needed to allow teachers to articulate their practice and create use cases or narratives. These can then be used by system developers to create the kinds of accessible, useable tools needed.

The ACETS Project [13] (Assembles, Catalogue, Exemplify, Test and Share) was funded by the JISC under the Exchange for Learning (X4L) Programme [14] . It devised a methodology for documenting and evaluating the use and implementation of reusable learning objects in the context of healthcare educational settings. Templates created to document teaching practice were based on IMSLD, but without explicit reference to the specification. More specifically, the project adapted a "natural" version of the IMS LD language in order to describe activities for the subsequent analysis of pedagogy in practitioner use of electronic learning resources. One of several JISC funded ELF (The E-Learning Framework)[15] reference model initiatives, LADIE [16] (Learning Activity Design in Education), it has developed a Learning Activity Reference Model or LARM (See fig. 1) to support "Learning Activity Authoring (the design and construction of learning activities and the discovery, specification, sequencing and packaging of content) and Learning Activity Realisation (the construction of the environment in which learning activities are to take place and execution of the learning activities themselves)"[17]. One of its main goals was to enable sharing and reuse of learning activity designs through the use of appropriate technologies and to open the way for the production of service based designs as opposed to learning designs that are restricted for use with a particular tool.

Fig 2

This project has mapped out required services for implementation in a demonstrator. Despite the problems inherent in sharing designs and services, the LADIE project has been able to produce an architecture for LARM. This has been possible through the creation of teacher generated use cases which identified a number of common requirements. The use cases were gathered using techniques derived from the [DialogPlus] toolkit [18] , which aimed to encourage teachers to model their pedagogic practice and to identifying emerging patterns.

The UNFOLD project has served to focus attention on a range of IMS LD related tools in development including the likes of Collage [19] , a project and a tool designed to capture flow patterns for computer based collaborative learning that can then be used to guide users on the best pattern for their needs. In a similar vein, UNFOLD also provided a platform for discussions on the creation and use of patterns on which to build templates that would capture best practice in e-learning and form the basis for a set of requirements to help in the development of suitable tools. Furthermore, the patterns also served as a map of best practice which could be used to inform the construction of IMS LD compatible Units of Learning by designers.

4. Relationship with services/components

The twin concepts of reuse and interoperability have received considerable backing in recent years among those who recognize their potential to reduce the amount of time and money spent on creating e-learning content and learning designs. In the case of the latter, IMS LD emerged in response to the need to structure and manage pedagogically complex activity structures which can subsequently be made available for reuse in a different academic context. However, the extent to which an intricate pedagogical structure can be modeled is contingent on the development of user-friendly tools and services. Thanks to initiatives such as the UNFOLD project and the e-Framework [20] , interest in the specification has promoted the development of the necessary types of tool which come in two flavours; general purpose tools needed to ensure the functionality of the Valkenburg Group [21] Architecture and more specialized tools needed to build and deliver Units of Learning. Examples of open source tools (general and specific) including Coppercore [22] , RELOAD [23] , Eduplone [24] and MOT+ [25] ,all of which are available as stabilized downloads today. A number of commercial implementations were also developed including, Perot Systems Edubox [26] , LAMS, eLive LD [27] suite, and GTK Press [28] .

In spite of the availability of such tools, their integration with IMS LD is limited at present due to a lack of requisite interfaces to enable the use of interoperable communications and collaboration services. By the same token, IMS LD currently supports only a limited range of services (email, discussion forums and a search facility) as a corollary of the need to ensure maximum interoperability of the specification during its earlier stages of implementation. One proposed solution to these shortcomings comes in the form of a Service Oriented-Architecture, supported by the e-framework which would obviate the need to have services dependent on the Coppercore runtime engine. In effect, the service approach would enable an environment to accommodate a learning design by invoking specific instances of services as and when required through the agency of improved communication between the player and the runtime engine. This separation of services and the runtime engine would be partly achieved by the advent of the JISC funded Service-based Learning Design system (SLeD)[29] project which has produced a service based Learning Design player. The SLED project is also credited with enhancing Coppercore in ways that would improve functionality in Reload such that it s UoL (unit of learning) testing and validation services shares the same communication path as [CopperCore].

Further additions to IMS LD are planned in relation to a more sophisticated form of assessment through the QTI specification. Research and technology developments planned as part of the TENCompetence project [30] aims to extend and integrate the recently released QTI version 2.0 with IMS LD. In effect, the project supports a raft of research and development initiatives including production of a formal specification intended to join communication & collaboration services to [CopperCore] and other such IMS LD runtime engines.

Despite being in the early stages of integrating web services with LD tools, the specification can be seen as an orchestration method for those web services it does include such as forums etc. as well as those that it will come to include over time.

5. Stakeholders concerns/issues (technical, semantical)

Throughout this chapter, key stakeholders have been identified in relation to this area. The OUNL and CETIS (representing the JISC in the UK) were the driving forces behind creating the IMS Learning Design Specification. As noted earlier the forerunner to the IMS Specification was EML, which was developed by the OUNL. The OUNL has committed to develop and deliver all its courses using the specification and is particularly interested in developing learning design within the area of competency based learning. The UK educational sector (from the school to University level) has also shown a commitment to investigating the use of IMS LD and the wider pedagogical implications of designing for learning, particularly within the e-Learning context. This work has been part of the pedagogy strand with the JISC's E-Learning Programme [31] . A two year Design for Learning Programme [32] was launched in May 2006. Twelve projects have been funded over two years covering the following themes; models of practice, pedagogic planning, implementation and technical developments. This programme has also funded projects looking at the use of LAMS in the classroom and the current state of practitioner tools. The UK [DfES] (Department for Education and Skills) [33] has also funded trials into the use of LAMS within the schools sector and has committed funding for schools to use LAMS.

There does seem to be a distinctly European commitment to learning design and the IMS LD specification. As noted the UNFOLD project provided a forum for developers, teachers and learning technologists to come together and move forward adoption of the specification. The original project was able to centralise and galvanise interest in IMS Learning Design and bring together many projects who, until that point, had been working in isolation. The UNFOLD community was also able to move forward thinking about the implementation and realisation of the IMS specification, mostly notably with the production of The Berlin Model [34] , which presents a new way of thinking about how web services can be integrated within a learning design runtime environment. The project was originally funded for two years (until December 2005); however it has received an extension to continue, albeit to a limited extent.

A number of the key players involved in the UNFOLD project (including OUNL, CETIS and Universitat Pompeu Fabra) are now collaborating again in the TENCompetencies project. The primary focus of this project is competency based learning, but there are two work packages within the overall project plan which are investigating the use of IMS LD within this context.

LAMS also continue to play a significant role in this field. The LAMS Foundation [35] is a not-for-profit organisation which manages the development and implementation of the LAMS system. The LAMS Foundation has also created an active online community providing support to users and also an area where LAMS activities can be shared and stored.

Technical

Since the release of the IMS Specification there has been one overarching concern - the lack of IMS LD specific tools, both to create and run compliant learning designs. Creating specification compliant, user-friendly authoring tools has proved to be very challenging. The difficulty has been in producing an environment that is accessible, which conceals the intricacies of the xml underlying the specification and doesn't require an in-depth knowledge of the specification from an end user. What is lacking is any kind of WYSIWYG (what you see is what you get) tool, which is common place in web authoring, allowing users with little or no knowledge of HTML to create web pages.

The release of the RELOAD open source editor has provided a breakthrough, in much the same way it did for the use of IMS Content Packaging, and indeed a number of projects using it as a basis to develop interfaces using wizard style templates (e.g. the Collage Project). When the specification was released no existing VLEs were able to run learning designs. Coppercore©, developed through the OUNL is still the only IMS LD runtime player. It has been adapted for the RELOAD player and other instantiations such as the SLED[36] player developed by the Open University[37], UK. Scalability is an issue for all LD players and work is ongoing both notably within OUNL community and through projects in the JISC Designing for Learning programme to address this issue.

In contrast, LAMS opted to take a learning design 'inspired' approach to the development of its interface. One of the recognised strengths of this system is its user friendly interface, which comprises of commonly understood drag and drop functionality to create and link components to create a learning activity. However the LAMS system cannot import an IMS Learning Design so is not a fully interoperable system.
Interoperability of learning design systems is still at the early stages of the specification lifecycle and is discussed in more detail in section 7.

Semantical

The growing interest in learning design has highlighted the need for commonly understood and shared pedagogic vocabularies to allow people and system to store, share and describing their learning activities. The recent JISC Pedagogical Vocabularies Scoping study [38] outlines the main semantical need for and barriers to the creation and use of pedagogical vocabularies within a learning design context.
It is possible to share and exchange text-based learning designs based on all kinds of more or less formal conventions. However, the point of the IMS LD is for designs to be interoperable across different platforms, and hence reusable and shareable. Any vocabularies used to support the specification must therefore:

Focus on learning activities, rather than on broad approaches to or theories of learning;
Separately identify and articulate (at least) the following elements:
- Type of learning activity
- Desired learning outcomes
- Learning systems or services required in the activity
- Other aspects of the learning environment
- Roles of participants in the learning activity
Reflect common usage among those educational practitioners who are likely to be developing and exchanging learning designs.

While some learning design projects have used or adapted existing pedagogic vocabularies, others have set out to develop and test their own. One reason for this seems to be the difficulty of identifying existing vocabularies that meet the requirements above. The elements of the IMS Learning Design specification, and particularly the focus on learning activity as the basic unit of description, already imply a logical structure to the description of learning that differs from the rationale behind many existing vocabularies. (Currier , Campbell, 2005)

Practitioners' (and indeed the needs of other stakeholders) needs for pedagogical vocabularies are clearly broader than the need for good quality descriptions or metadata. The discussions that have taken place (both since the inception of IMS LD and obviously previous to that time) have highlighted the fact that educational practitioners and learning technologists perceive a real and pressing need for pedagogical vocabularies. There is a need to find a way to describe common practice so that resources can be discovered and shared across the community. Although there are many models of practice in existence in education, the models used by those working in the field of e-learning do not necessarily relate to the models used by teaching practitioners. Many teachers are not reflective practitioners, and so do not describe their teaching approaches with the formal terms used by educational researchers. Often, teachers are not aware of these formal terms and models, and the approaches that they personally adopt are implicit. As described in section 3 there is growing interest in the use of patterns and pattern languages to help describe generic learning activities.

Comment from Steve. Is it a little a harsh to say that most teachers are not reflective practitioners? Their analysis is based on a different set of empirical perspectives.

6. Interoperability scenarios

Although interoperability is a fundamental part of the IMS LD, as discussed earlier, there is currently a lack of tools both for authoring and then playing designs at 'runtime'. An overview of tools identified through the UNFOLD project are shown in the following tables.

Learning Design editors

Higher level editors designed to facilitate the process of XML coding currently include:

Name	Producer	Purpose	IMS LD Level	Ownership
MOT+	LICEF, Université de Quebec	General purpose graphical editor	A,B and C	Open Source
eLive	eLive GmbH Germany	General purpose graphical editor	A and B	Proprietary
[EduPlone] Sequencer	[EduPlone]	specialised template based editor	A	Open Source
[EduCreator] Editor	Chronotech	General purpose editor	A	Proprietary
LAMS -	LAMS Foundation	Learning Activity Management System with IMS LD Level A export NOTE: Not fully compliant with IMS-LD	A	Open source
COLLAGE	University of Valladolid	High-level specialized Learning Design authoring tool for collaborative learning	A	Open Source

Examples of low level editors affording technicians direct access to the code

Name	Producer	Purpose	IMS LD Level	Ownership
aLFanet	aL.Fanet project	General purpose tree editor	A	Open source
ASK-LDT Editor	EU project ICLASS: Informatics and Telematics Institute. (CERTH) Greece	Graphical editor.	A, B	Freeware
[CopperAuthor]	Open University of the Netherlands	Tree based editor.	A	Open source
Komposer	GTK Press	Tree based editor	A	Proprietary
COSMOS Editor	Yongwu Miao, University of Duisburg	General Tree based editor	A, B and C	Open source
RELOAD Learning Design Editor	RELOAD Project (JISC)	General Purpose tree based editor	A, B and C	Open Source

Runtime players

Name	Producer	Purpose	IMS LD Level	Ownership
Coppercore Learning Design Engine	OUNL	Core of Learning Design player	A, B and C	Open source
RELOAD Player	RELOAD project	Viewer for integration with RELOAD editor	A,B and C	Open Source
SLED Player	OU UK and OUNL	Service based LD player built on [CopperCore]	A,B and C	Open source	There are also a growing number of tools which are not full editors or players such as the [DialogPlus] Toolkit. Primarily a pedagogic planning tool, it can export documents as IMS LD fragments. One of the key issues facing the learning design community is the issue of the integration of runtime players into existing learning systems. It could be argued that unless there is mainstream availability of runtime systems then the IMS LD specification is of no value as no systems can run it. The LAMS community have been active in beginning to establish links with a number of learning systems including Blackboard [39] and the open source Moodle [40] Community. Concurrently there are the beginnings of moves towards IMS LD integration from other systems including Boddington [41] and Moodle. Another key issue to be addressed by the community is how to adapt current digital repositories to allow effective storage and sharing of learning designs.

7. Relationships with the adoption life cycle

Olivier's (2003) representation of the adoption life cycle of a specification is shown in Figure 2.

Fig 2

Fig 3
In terms of IMS LD, at the time of writing, the shaded area in Figure 2 reflects the status of the community. We are now at a stage where we have the first iteration of a specification with at least one open source reference implementation. On the technical side, we are beginning to see the implementation of runtime systems and developers are working towards establishing interoperability. On the pedagogical side, we are now developing a small (but growing) group of practitioners who are creating the learning designs which can be run by these systems and accessed by teachers and learners.
In conclusion, in terms of tools, systems and actual learning content that could be used in a real teaching scenario, we are very much at the stages of first generation tools, activity development and services. However, the awareness raising and support for effective practice shown in the topmost box on the right hand side of the diagram, is already advanced and being put into place through the work of OUNL, JISC/CETIS UNFOLD, LAMS and other key stakeholders as identified in Section 5. Critical mass of interoperable systems, services, content exchange and delivery has not been achieved yet.

References

Bailey, W. (2005) What IS... IMS Simple Sequencing? http://www.cetis.ac.uk/lib/media/ss2brief.pdf
Currier, S and Campbell L (2005) JISC Pedagogical Vocabularies Review http://www.jisc.ac.uk/elp_vocabularies.html
Currier, S and Jeffrey, A. (2005) What Is... IMS Learning Design? http://www.cetis.ac.uk/lib/media/WhatIsLD2_web.pdf
Koper, R. and Tattersall, C. (2004) EML and IMS Learning Design: from LO to LA. Heerlen: OUNL. http://dspace.learningnetworks.org.
Koper, R and Olivier, B. (2004). Representing the Learning Design of Units of Learning. Educational Technology & Society, 7 (3), 97-111
Kraan, W (2003) The teachers teach the techies
http://www.cetis.ac.uk/content/20030415235439
[McAndrew], P. (2004) (The Effectiveness of Resources, Tools and Support Services used by Practitioners in Designing and Delivering E-Learning Activitieshttp://www.jisc.ac.uk/uploaded_documents/practioner-patterns-v2.doc
Olivier, B (2003) Reaping the Benefits of e-learning standards http://www.eun.org/etb/95_B_Olivier_Adoption__eLearning_Standards.pdf

[1] http://www.imsproject.org/learningdesign/index.html
[2] http://www.ou.nl/eCache/DEF/36.html
[3] http://learningnetworks.org/forums/index.php
[4] http://www.cetis.ac.uk
[5] http://www.unfold-project.net
[6] http://www.imsglobal.org/content/packaging/index.html
[7] http://www.imsglobal.org/question/index.html
[8] http://www.imsglobal.org/simplesequencing/index.html
[9] http://www.adlnet.gov/scorm/index.cfm
[10] http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf
[11] http://www.lamsinternational.com/
[12] http://www.jisc.ac.uk
[13] http://www.acets.ac.uk
[14] http://www.jisc.ac.uk/programme_x4l.html
[15] http://elframework.org
[16] http://elframework.org/refmodels/ladie
[17] http://www.jisc.ac.uk/index.cfm?name=elfref_southhampton
[18] http://www.dialogplus.org/
[19] http://gsic.tel.uva.es/collage
[20] http://www.e-framework.org/
[21] http://www.valkenburggroup.org/valkenburggroup-org.htm
[22] http://coppercore.sourceforge.net/
[23] http://www.reload.ac.uk
[24] http://eduplone.net/
[25] http://www.licef.teluq.uquebec.ca/francais/real/demot.htm
[26] http://elearning.surf.nl/edubox/home
[27] http://www.elive-ld.com/content/index_eng.html
[28] http://www.gtkpress.com/default.aspx?pid=m_1013
[29] http://www.elframework.org/projects/sled
[30] http://www.tencompetence.org
[31] http://www.jisc.ac.uk/programme_elearning.html
[32] http://www.jisc.ac.uk/index.cfm?name=elp_designlearn
[33] http://www.dfes.gov.uk/
[34] http://dspace.ou.nl/bitstream/1820/505/6/architecture_berlin+model+2.pdf
[35] http://www.lamsfoundation.org
[36] http://www.jisc.ac.uk/index.cfm?name=elfdemo_livhope
[37] http://www.open.ac.uk/
[38] http://www.jisc.ac.uk/elp_vocabularies.html
[39] http://www.ble.ac.uk/
[40] http://www.moodle.org
[41] http://bodington.org/index.php