Skip to content.
Sections
Personal tools
You are here: Home » Features » The Reference Model Approach - using services to meet our needs

The Reference Model Approach - using services to meet our needs

Sarah Holyfield
Last modified 06 Oct, 2006
Published 12 Jul, 2006
This article explores the role reference models are playing in the e-Framework, and describes the projects that have been developing this approach.

The e-Learning Programme has been developing technical tools, investigating the use of technology from a pedagogical point of view, and exploring the role it can play in supporting lifelong learning. The intention has been to ensure that the tools developed are able to support a wide range of pedagogical approaches, and to provide flexibility in how different types of educational organisation can adopt systems that will be suited to their particular needs.

Over recent years colleges and universities have implemented large systems to provide their IT infrastructure, but these systems were not originally designed to easily share and exchange data with others, and this has become an increasingly difficult problem as the trend is towards a pattern involving the learner moving between several institutions, the workplace, and sometimes between countries.

This is also a more general and well-recognised problem in the business world, and technologies have been evolving to meet this challenge. The new software design paradigm focuses on the use of small components which can easily exchange data with each other and which can be ‘mixed and matched’ in what is known as a service oriented approach.[1]

Whilst this approach means that it is theoretically possible to create an ‘ideal’ system that will do exactly what you want, the challenge then becomes one of developing a good understanding at a ‘human level’ of what activity you want technology to support (eg what pedagogical approach you wish to take, or what administrative task/headache you want technology to help with), and at a technical level of what tools are available, how they behave, and how they can be combined. This is a classic ‘bridging problem’.

The e-Learning Programme and the e-Framework

The work that has taken place in the e-Learning Programme has been addressing this challenging area by considering how a service oriented approach can be taken in education. It initially produced a framework of potential services which were of specific relevance to e-learning (ELF), but this has more recently been expanded to incorporate the field of Research, and a collaboration with international partners who have been thinking in the same way has developed. This has resulted in the-Framework initiative [2]

The aim of the UK/Australian joint activity is to produce an evolving and sustainable, open standards based, service-oriented technical framework to support the education and research communities

The next stage of thinking has been concerned with deciding which services might comprise the e-Framework, elaborating the detail of these, how they behave and work together, and how they might be combined to meet commonly experienced problems or tasks. As a result of this work, the key concept of the ‘reference model’ has emerged. [3] (see below for discussion about new termonology for these concepts)

Reference models

One of the key problems in designing software is the difficulty of specifying what it is that users actually want to be able to do (user requirements). Programmers need this information to be articulated in a very precise way in order to code software, yet users may not even know what is possible with technology so may not be able to envisage the role it could play. Developers may not have a good understanding of the professional area the users are involved in, so they may not be able to help them to explore what is possible, or feasible. There is therefore a need to establish a meaningful dialogue between users and technologists.

The idea underpinning the reference model is that it should bridge these two worlds. It includes a practitioner part and a technical part and ‘is a way of ensuring user involvement in the development process’ [4] This involves creating a model at a number of levels thus allowing the needs of the user to be expressed in a rich way and the developer’s need for technical precision to be met.

The structure of a reference model is being developed and includes several levels which move through from a ‘human level’ view to a ‘developer level’ view [5]


Reference Models set out:

  • the learning, teaching or research problem addressed
  • a set of tasks needed to fulfil this
  • the human and computer based workflows
  • the agents, applications and/or tools used
  • the data flows and operations involved
  • the services that will be called on
    • their ‘orchestration’ (how several serve a single user)
    • and their ‘choreography’ (how several participants carry out common workflows using services)
    • the service interface specifications and any profiles (variations) to be used, with references to the definitions of the services used

A “Reference Model” bridges :

  • the World of Users and their Work
  • the underlying Technical Services and their associated Specs (The Wall)


The following diagram [6] shows how the reference model allows user needs (related to a particular problem) to be met through a combination of services. A design can then be created based on this information leading to an implementation which can then be referred to by others who are trying to solve the same problem.

ref model slide one3

(For a more in depth discussion see the Briefing paper on the e-Framework and Service-Oriented Architecture [7] )

JISC reference model projects

Several projects have been funded by JISC to explore how best to develop the concept of the reference model. Projects were asked to focus on a number of domain areas including e-assessment, learning content, personal development planning, enterprise, and personal learning environments. These were aligned with several CETIS Special Interest Groups (SIGs) [8]. These domain areas vary widely from those with an enterprise focus, to the more complex and less easily modelled areas within the pedagogical arena such as e-assessment, and they also vary in maturity. These factors meant that the projects could play a very valuable role in developing the concept of the reference model.

The call [9] in January 2005 invited proposals to specify reference models for the e-learning framework. Projects will be expected to take an evolutionary approach to the development of the reference models through iterative cycles of development and implementation

The first part of the project was to define the scope of the domain area chosen, create a picture of what is currently happening in this area and what gaps exist, and map this information to the services already defined in the e-Framework [10]. They were asked to then provide a ‘common solution pattern’ for the application domain, develop use cases within it and define which e-Framework service definitions would best support this pattern.

The second part of the project was required to develop a reference model implementation which would support the development of the model, or produce a proof of concept.

Six projects were selected and provide a useful variety and range in terms of the scale, precision and maturity of the domain they are addressing. A brief summary of the projects is in the box below and more detailed links are provided in the footnotes.


Reference model projects

Each reference model is a service-oriented description of a collection of use cases, workflows, service descriptions, toolkits and other artefacts specific to the issues of a particular problem domain. The intention is that each reference model will enable new work within the field to build on a service-oriented view of the domain and ensure that new projects are aligned with the strategic direction of the e-Framework.

There are six reference model projects, addressing the domains of:

  • Assessment (FREMA) [11]
  • Learning Activity Management (LADIE) [12]
  • Course Information Exchange (XCRI) [13]
  • Course Validation (COVARM) [14]
  • e-Portfolios (EP4LL) [15]
  • Personal Learning Environments (PLE) [16]

A service-oriented view of the assessment domain is explored in the e-Learning Framework Reference Model for Assessment (FREMA) and similar perspective of the domain of learning design and learning activity management is articulated in the Learning Activity Reference Model (LADIE).

The Exchanging Course Related Information (XCRI) reference model has pioneered an XML schema for representing academic courses. The XCRI work has strong associations with reference models that take a service-oriented view of the course validation and student progression: the Course Validation Reference Model (COVARM) and the ePortfolio Reference Model (EP4LL)

The reference model for Personal Learning Environment (PLE), like Web 2.0 approaches and service mash-ups, explores the possibilities of a persistent environment for life-long learners that can interact with institutional systems and services. It also looks for the novel service synergies and combinations that arise when academic resources are factored as discrete services.

There is a strong expectation that the outputs of each of the reference models will inform the direction of new projects to ensure alignment with the principles and practice of service-orientation within the e-Framework.

[17]

These projects began work in March 2005 for 12 months and have been given extension funding until the summer of 2006 to complete their work and incorporate the reference models into the e-Framework

The JISC-CETIS Conference in November 2005 [18] considered the role of reference models in some depth, both at a domain level but also at the general level of the e-Framework. Many more reference models will be needed, and will evolve, and there will be overlaps and synergies. Dan Rehak described many reference models ‘floating above a sea of services’ [19]

sea of services

Issues emerging

During the life of the projects there have been many meetings and discussions, and the following points represent a summary of some of these.

Sustainability

The long term life of technical development work is always a serious concern [20], and a recent meeting explored how to develop the e-framework services, to update the service descriptions and definitions and decide who would be looking after which services [21]. An active user and developer community will be required to maintain and develop the reference models. Other factors such as scale and stability will be important, as will the relationship between generic patterns and specific implementations. The challenge for institutions will be ‘adapting their processes to a changing world whilst maintaining enough stability - the vision of service oriented architecture’ [22].

Shared understandings

All professional and academic disciplines necessarily develop precise terminology – or ‘jargon’ to those outside the field – yet reference models are concerned with enabling a dialogue to take place between very different groups. There is therefore a need for a shared language, along with an agreed set of definitions ranging from the descriptive to the technical, and this work is being undertaken by the e-Framework.


nb These definitions will lead to some changes and the term reference model will be replaced by Service Usage Model in the future.

The community- ownership and control

The projects have all involved their domain area communities through the CETIS SIGs, and the JISC-CETIS conference, as well as other channels. They have also often worked closely together and this raises the question of whether there should be a special interest group for this sort of work?

The long term life of these models requires community ownership, yet this involves a tension between the central control required for the e-Framework to function, and the need for local implementation and development which will help in the validation of the models. The involvement of vendors is also an essential dimension.

What next?

This work raises some important questions. It is well understood that the adoption of technology requires the user to examine what it is they wish to do and why, and the role that technology could play in this - an essential element of developing a reference model. This should, however, also apply at the more general level of the system as a whole and what the service oriented approach means for education, and how we organise ourselves.

There are implications at ethical levels as well as technical. The way a learner is perceived will determine how technological systems are implemented – is the learner at the centre taking control of their learning space over their lifetime, or is the focus on the centralisation of information from the point of view of the administration? Pragmatically, the answer is likely to be that both of these requirements should be met, but the question of ownership of personal information will need to be addressed. These are strategic and policy considerations, and will determine how funding is channelled into further development. The reference model projects can play a significant role in informing these decisions.

What has been striking about the reference model projects has been the level of involvement and commitment in what has been a very intense process. There have been, as expected and hoped, many perspectives on the evolving concept of a reference model, and those involved have themselves bridged the communities of e-learning practitioners, developers, academic computer scientists and many others.

The range of domain models means that the projects have explored established areas alongside the new, and e-administration areas alongside the pedagogical and learner centred, and they have begun to join these up together. The question is whether the e-Framework can now enable a bigger picture with a sufficient level of flexibility to emerge?

________________________________

References

[1] A non- technical guide to technical frameworks - Part One

[2] e-Framework website

[3] Reference Models the next important step introduced the idea of reference models in a general way and explained some of the concepts involved, and readers may wish to look at this before reading on.

[4] Bill Olivier in conversation

[5] The e-Framework presentation by Bill Olivier, Technical Director, Systems and Technology, JISC, 2006

[6] as above

[7]Briefing paper on the e-Framework and Service-Oriented Architecture

[8]CETIS

[9]Circular 10/04: Specification of e-Learning Framework Reference Models

[10]see the e-Learning Framework for the view at the time

[11] Resources for FREMA - e-learning Framework REference Model for Assessment conference

JISC-CETIS Conference 2005, Assessment Theme

FREMA website

The JISC E-Framework Reference Models

FREMA briefing document

Presentation - FREMA

[12] Resources for LADIE Learning Activity Design in Education

LADIE outputs

JISC-CETIS Conference 2005 - Learning Design theme

LADIE on the ELF website

[13] Resources for XCRI eXchanging Course-Related Information

JISC-CETIS Conference 2005 - Institution Level Systems Integration Theme

The JISC E-Framework Reference Models

XCRI briefing sept 05

XCRI presentation

[14] Resources for COVARM Course Validation Reference Mode

COVARM website

The JISC E-Framework Reference Models

Managing quality and improving efficiency in the course validation process

JISC-CETIS Conference 2005 - Institution Level Systems Integration Theme

COVARM presentation

[15] Resources for EP4LL ePortfolio for Lifelong Learning

ePortfolio for Lifelong Learning on ELF website

JISC-CETIS Conference 2005 e-Portfolio Theme

[16] Resources for PLE Personal Learning Environments

JISC-CETIS Conference 2005 Personal Learning Environment (PLE) theme

[17] Briefing paper on the e-Framework and Service-Oriented Architecture

[18] JISC-CETIS Conference Programme Keynote presentations

[19] Presentation on Reference Models by Daniel R. Rehak

[20] Tackling Sustainability

[21] Adopt a Brick!

[22] Quotation from Bill Olivier at Capital Programme Town Meeting 10th may 06

 

Supported by JISC Supported by CETIS
Powered by Plone