Skip to content.
Sections
Personal tools
You are here: Home » Features » Reference Model/Domain map projects: Surfacing complexity

Reference Model/Domain map projects: Surfacing complexity

Christina Smart
Last modified 07 Nov, 2006
Published 30 Oct, 2006
For the last 18 months a group of six JISC funded projects have been mapping and modelling a number of e-learning domains to provide a link between the e-Framework for Education and Research services and users. On the 10th October these projects gathered at Aston University to share and reflect on those experiences and to pass on their wisdom to those wishing to bid into the next round of funding.

We have discussed the JISC Reference Model (now Domain Map) projects previously on this site and the role they have played in bridging the worlds of users and developers thus enabling the development of interoperable e-learning systems that both address user needs as well as system constraints [1].

At the same time the projects also bridge the space between the high level e-Framework for Education and Research and the practical needs of specific users and communities [2].

Here is how JISC currently describes these projects:

The term "Reference Model" is currently used in the e-Framework as a place holder for the concept of developing models based on learning, teaching, research or business requirements to show how one or more Services can be used to meet the described need. However, it has subsequently been agreed that "Reference Models" was not a good term to describe the range of outputs resulting from the set of projects commissioned, nor what the e-Framework was asking the community to contribute to the initiative. As a result, the concept of "domain maps" has emerged, which consist of a set of models (functional, practice, process, information and service usage models). Domain Maps play a key role in bridging the world of users with the underlying invisible world of services. They thus also provide a route for institutional planners and users to finding appropriate tools and services that meet their needs.[3]

The first set of JISC funded Reference Model/Domain Map projects started work in March 2005 when the e-Framework was evolving from the e-Learning Framework (ELF) and thinking in this area was changing daily.

The projects covered the following domains:

  • Course Validation, COVARM
  • Course description, XCRI
  • E-portfolios, Ep4LL
  • Learning activity design, LADIE
  • Assessment, FREMA
  • Personal learning environments, PLE

This meeting in Birmingham was therefore the finale for the current projects, an opportunity to discuss findings and to inform the next round of projects that will emerge from the capital programme call.

Purpose and Aspirations

Bill Olivier head of technical development at JISC kicked off the day by outlining the purpose of the Domain Map project [4]. He asked the question; why is it so difficult to specify requirements? In reality, it is actually hard to co-design software and new ways of working, users don’t know what is feasible and technologists don’t understand user practices. In order to bridge the world of users and the underlying technical services, we need to make clear the human context and the machine context. The Reference Model/ Domain map projects did this by developing a number of elements:

  • A human domain map or ontology
  • Workflows of processes and practises – both as is and what could be
  • The information models and flows. Process models describe human level processes, Service usage models – SUMs describe machine level processes.
  • An outline of services needed, the user interface, services and a co-ordination layer which orchestrates how the services interact.

At this meeting each project gave a 30 minute overview of their work and were encouraged by programme manager Heather Williamson start with a summary of their work in a single sentence.

COVARM – Course Validation Reference Model

Balbir Barn summarised the aim of the COVARM project in the following sentence:

“To deliver a UML model based software application of the course validation problem domain from analysis process and information models to executable business processes.”

AND

“to provide generic capabilities for addressing other problem domains in the HE admin sector by providing a supported method.”

Course validation is a process at the very core of HE institutions. It’s a knowledge intensive process by which new courses are developed and quality assured by the sector. The project, a collaboration between the universities of Thames Valley, Manchester, Manchester Metropolitan and Staffordshire, and began by looking at the context in which course validation takes place. Universities develop new courses to attract students from new markets (e.g. overseas students), to adapt to changes in business and culture, to offer courses which are more flexible for lifelong learners, and to differentiate themselves from other institutions, in other words to be competitive.

As a process course validation is very knowledge intensive relying on the collaboration between a number of experts. The project took a business process based approach to analysing the problem space and developed four case studies of the course validation processes in different institutions. For each of the case studies a process model was defined. These process models generated significant interest in the host case study organizations. These processes were synthesized into a single canonical reference process which was then used in subsequent development. This process required both computer and human interaction.

COVARM went on to develop two technical scenarios for the canonical process: (i) the preparation of a proposal for a new programme and (ii) the organising of a validation event. The scenarios were implemented as WSDL services and choreographed using BPEL process definition technology. To support the development process The team have developed methodology collateral that can be applied to other similar e-admin problem spaces, using the COVARM_PDK tool to support developers implementing business process led Applications [5], [6].

eP4LL e-Portfolios for Lifelong Learning

Angela Smallwood, Peter Rees Jones and Sandra Kingston gave an overview of the e-portfolio domain map project. In a sentence:

“The reference model has mapped the HE admissions process concentrating on a Personal Profiling Service, but scoping out Advice and Guidance, Feedback and Personal Planning Services.”

It is now widely recognised that the domain covers a number of different types of record including e-portfolios for learning, assessment, presentation and transition from one institution to another. At the beginning of the project it was recognised that transition e-portfolios were the most mature of these types. The project has worked with UCAS and focussed on the area of HE admissions and were guided by the Schwartz review into admissions to higher education which emphasises that the application procedure could be enhanced by technology without forgetting the needs of the individual [7].

For this reason the project has worked in detail on a Personal Profiling service as well as scoping out Advice and Guidance and Personal Planning services. Their analysis while focussed on the HE admissions process could also be applied to Human Resources processes and the process of applying for a job [8],[9].

Use case for the regional partnership

Use case for the regional partnership

The project proposes a thin (or slender) e-portfolio – where applicants pull in supporting information from previous e-portfolios without maintaining their information centrally. Deliverables from the project include an outline domain map, a re-factoring of regional partnership interactions and a thin model for an e-portfolio. Prototype web services are now being developed by two vendors.

FREMA e-Learning Framework Reference Model for Assessment

Dave Millard and Yvonne Howard summarised the FREMA project in the following sentence:

“A Community developed Reference Model for Assessment.”

The assessment domain is a broad one – covering everything from full online exams to technology assisted assessments and the project team described it as a “brown field” site – because there have been many developments in this area over the years and the majority of institutions – if not all institutions will be engaged in some sort of e-assessment activities.

They also took a layered approach beginning with the broad domain map and working up through service profiles to the most concrete but focussed deliverable being the reference implementation.

Being such a broad domain the team decided to take the approach of putting a semantic wiki in place to capture the breadth of activity. All entries are tagged – also can be searched allowing potential developers to identify gaps that are ripe for development – which are well defined – and have appropriate specifications but no supporting software. In addition the project has also produced concept maps of e-assessment processes and entities to help users navigate within the wiki.

In developing service usage models the project took an agile approach and worked up scenarios, usage models and use cases and service profiles in the area of summative assessments. But the team emphasized again the complexity of processes that different institutions employ to do summative assessment and the team didn’t want to be prescriptive about how institutions might want to build their own systems. For this reason the project defined services in an abstract way, focusing on what services should exist, and how they could collaborate to fulfil a use case, rather than on defining concrete data models and methods [10], [11].

LADIE - Learning Activity Design in Education

This project, a collaboration between Glasgow Caledonian, Dundee, Southampton Universities and Intrallect, was represented by Isobel Falconer. It aimed:

“To develop a learning activity reference model firmly based in practical experience of teaching and learning.”

The project used Beetham’s definition (2004) of a learning activity [12]:

“An interaction between a learner or learners and an environment that is carried out in response to a task with an intended learning outcome.”

The project has addressed the broad area of Learning Activities from a top down and bottom up perspective. At the top level the team analysed user needs through a series of practitioner workshops from which they developed 16 use cases. One use case was based on a 1st year journalism class which included elements of presentation, discussion (forum), brainstorming (concept mapping), testing and a research element.

At the bottom end they looked at services and developed a LARM (Learning Activity Reference Model) architecture which elaborates the services required for authoring and delivery of learning activities. The model identified areas of overlap in the e-portfolio, assessment and repository areas domains.

The project also developed three guides -

  • A Pedagogy Guide – for practitioners,
  • An Implementation Guide – for learning technologists, and
  • A Services Guides for software developers.

A number of issues emerged including that of granularity, how big or small is a learning activity, is it a whole course or is it reading a book?

One requirement that users frequently requested was the need to be able to change and adapt sequences as they are running.

Another recurring tension was that between the need to capture current practise and the potential for future practise – teachers are constantly revising, reflecting and improving their practise so will inevitably focus on what they could do next [13], [14].

XCRI -Exchanging Course-Related Information

The XCRI project differs from the other projects in that it came from its community. Mark Stubbs chartered the progress of the project which started with a recognition by some members of the CETIS Enterprise SIG that while specifications existed for exchanging information about people, groups and membership – there was no standard way of exchanging information about course descriptions a core part of the business of education.

XCRI’s first task was therefore to address this gap by developing an XML schema which reflects the needs of many institutions.

The XCRI schema was designed for use in the area of course management, where institutions need to keep track of different versions course information, that might be held in a number of databases across an institution. The schema can also be used for course advertising – on web sites and prospectuses When version 1 of the schema was released many projects it was clear that many institutions were interested in the schema for course advertising purposes.

In response the team have produced a CAP – course advertising profile - of XCRI a slim version of the schema for this purpose. In addition they have built an aggregator to link course information to Google maps information to help students find courses. In the next round of funding a number of mini projects will be exploring the use of the CAP and the aggregator [15],[16], [17].

Personal Learning Environments

Wilbert Kraan summarised the PLE domain map project as:

"An Exploration around learner owned toolsets for formal and informal learning."

The project came out of a recognition that VLEs are for the most part designed as systems for institutions not for individuals. Moreover every time someone enrols at a new institution they have a new system to learn, so VLEs don’t facilitate lifelong learning. The personal learning environment project acknowledged the following concepts:

  • Multiple concept learning: that people learn in a variety of ways
  • Formal and Informal; that people use a blend of formal and informal (non-institutional) learning
  • Role switching; that people often switch roles between teacher and learner
  • Self organisation; that people will organise their work and tools they need

Many of these concepts are associated with the Web 2.0 (social software) philosophy.

Like the other projects PLE took a top down, bottom up and exploratory approach. Top down was the scenario development and theory driven analysis based on the work of Heidegger, Winograd & Flores and Beer [18].

The project team used pattern analysis to functionally categorise current and emerging applications. Broadly technologies could be grouped into categories such as resource patterns, conversation, workflow, context and team.

The team produced a number of models to illustrate alternative ways that a user could interact with a number of service providers using their personal learning toolkits. The toolkits would interact through feeds and conduits:

A feed is a mechanism for enabling a personal learning toolkit to aggregate the objects managed by services such as resources and activities. A conduit is a mechanism for enabling the toolkit to share, create and affect objects managed by services. A service provider might provide different types of service for Resources and for Co-ordination of groups of learners with their toolkits.

How a personal learning toolkit interacts with a service provider.

How a personal learning toolkit (plt) interacts with a service provider.

In the exploratory part of the work two prototypes were produced, PLEX – a desktop application and a web based tool PLEWeb. Both tools allow users to organise activities, people and resources and aggregate feeds from various web services chosen by the user. One issue that arose from the prototype work was the difficulty of defining user needs with such a potentially generic tool.

The project raises fundamental questions about boundaries – what are the boundaries between work, study and our personal lives? It also, as one person pointed out, challenges the very core of what formal education is about [18],[19].

Emerging Issues

Various issues emerged during the projects including:

  • Granularity – what is the most useful level to describe services at? Too detailed and there is a danger of missing recurring patterns – too high level and you miss important complexities. The COVARM team quoted Einstein: “Everything should be as simple as possible, but no simpler.”
  • Domain specificities and language – it is important to use language and approaches that suit the domain. For example the FREMA project used the term ‘candidate’ a term widely used in the assessment domain.

Outputs and new projects

So how is the work of these first Reference Model projects being taken forward? Answer in two ways, firstly Phil Nicholls will be mining the project outputs and placing them on the e-Framework for Education and Research web site so that they are freely available to institutions and developers [20]. Secondly through a group of new projects in the current JISC Capital Call.

New projects in Capital Call

Projects funded through the e-learning strand of the current Capital Programme Call [17] will also build on the work of several of the original reference model projects in the following areas:

Assessment

In the area of technical developments to support teaching and learning there is a call for three projects to develop item authoring, item banking and assessment delivery open source tools that conform to the QTI 2.1 specification. A need for these tools has been identified by the FREMA project and the JISC-CETIS Assessment Special Interest Group.

Technology Supported Learning Environments

Four projects will build on the work of the Personal Learning Environments project the current call includes projects which will demonstrate the integration of user-owned technologies with institutionally owned educational systems.

Administration of learning and teaching

There are many opportunities to gain efficiencies in administrative processes through the use of technology and reference model projects in these areas demonstrated that many areas are now ready for development.

Admissions demonstrators

Between four and six projects will investigate and develop systems and processes to improve HE admissions in the areas of personal profiles, applicant feedback, and accreditation of prior experiential learning and e-portfolio admissions. This work will clearly take forward the work of the e-portfolio for lifelong learning project.

Course description and discovery

8-10 mini projects will implement and test the XCRI course information specification.

Course validation

Two projects will develop a system to support course validation based on the COVARM reference model.

New Domain Maps

There will also be between two and four projects to develop Domain Maps for key administrative functions including accreditation and achievement and student fee management.

Conclusion

What the reference model/domain model projects have sought to do is very difficult and complex – as Bill Olivier pointed out at the start of the day the projects have been trying to address some fundamental questions:

  • How do we model our world? - how can projects identify and work most effectively with their communities of practice? To identify critical problem areas and new opportunities?
  • How do we develop open service specifications?
  • How do we support adoption?- How should we be supporting institutions adopting a service oriented approach?- How can we share experiences and communicate successes (and failures) more widely?

It is clear that the Reference Model/Domain map projects have made a significant progress in answering these questions for their particular domains and developed many useful resources. They have also helped inform JISC’s thinking about the e-Framework for Education and Research which has evolved during this time.

New projects in the Capital Call will be building on the work of the projects in key areas – particularly in e-administration where focussed developments can bring real efficiency gains for the sector.

References

[1] Reference models - the next important step Sarah Holyfield, July 2005

[2] The Reference Model Approach - using services to meet our needs. Sarah Holyfield, July 2006

[3] JISC Definition of a Reference Model/Domain Map project

[4] Bill Olivier, The e-Framework from Reference Models to Domain Mapping 10-10-06

[5] COVARM project presentation 10-10-06

[6] COVARM web site

[7] Schwartz review Admissions to Higher Education in England

[8] ep4LL project presentation 10-10-06

[9] ep4LL project web site

[10] FREMA project presentation 10-10-06

[11] FREMA project web site

[12] Review: developing e-Learning Models for the JISC Practitioner Communities Beetham 2004

[13] LADIE project presentation 10-10-06

[14] LADIE project web site

[15] XCRI project presentation 10-10-06

[16] XCRI project web site

[17] JISC Capital Programme Call September 2006 (link to Appendix D e-Learning is at the bottom of the page)

[18] PLE project web site

[19] PLE project presentation 10-10-06

[20] e-Learning Framework for Education and Research

[21] Presentations from all the projects are available from: http://www.jisc.ac.uk/refmodelssept05.html links are at the bottom of the page.

 

Supported by JISC Supported by CETIS
Powered by Plone