This is a guest post by Andrew G. Booth and Brian P. Clark for On the Horizon series on distributed learning. Professor Andrew G. Booth, a National Teaching Fellow, is Professor of Online Learning in the Faculty of Biological Sciences in the University of Leeds where he has deployed the Bodington System to support online distance teaching, primarily of Bioinformatics, in partnership with the University of Manchester. Prior to this appointment, he was Director of the Flexible Learning Development Unit, where he headed the team that developed the Bodington System and was one of its original authors. Following a period as a comedy writer for radio and television, and then 10 years as a lecturer in physical chemistry at the University of Sunderland, Dr Brian P. Clark is now a developer/tutor for the online MSc Bioinformatics at the University of Leeds teaching modules in the bioinformatics of protein structure and molecular modeling.
Our position in the learning environment debate can be summarised by the following statement. As well as students being able to create their custom learning environments, we believe that it is important that teachers, too, are able to create their custom teaching environments.
The VLE has been a useful and effective device in university education and it is not yet time to abandon it. One viewpoint that we do share with many, though, is that the monolithic VLE doesn’t offer the flexibility required to support education in today’s fast-moving technical arena. We are currently developing a Service-Oriented Virtual Learning Environment (SOVLE) that is composed of a set of pluggable resources that have Web application front ends and Web service back ends. The latter can provide programmatic access to services using client-owned technology.
An important benefit delivered by the SOVLE is that constituent services are able to take part in teaching and learning workflows using established workflow technology such as Business Process Execution Language (BPEL) and Business Process Management Notation (BPMN). This approach has been dismissed in the past on objections relating to non-scalability, lack of security provision and an inability to handle human-based tasks well. None of these objections apply now.
In order for our SOVLE approach to be represented properly we need to revise some of the terminology being used in the discussions. Firstly, when one talks about centralisation it is best that this refers to the centre of the institution. In this case, we can distinguish between decentralisation, the distribution of services away from the management of a central IT service, and externalisation, where students use services that are not provided and managed by the university. With decentralisation, the locus of control is redistributed within the institution, and with externalisation, the locus of control of a service ends up outside of the institution. The process of disaggregation of a monolithic VLE would then hold no connotation of changes in control. Disaggregated services could reside with the centre, be decentralised, externalised, or replaced by an equivalent external service.
All this might seem like an irrelevance, especially to those who have already discarded Service-Oriented Architecture (SOA) as a possible future direction for higher education. However, given that the major body supporting IT development in the UK has adopted a service-oriented approach (little soa, just to confuse everyone) it is too soon to abandon this model despite the general lack of interest (and funding) in the research and development of a suitable platform for higher education: soa needs SOA. Should SOA be implemented in the universities it would almost certainly be necessary to decentralise the management of Web services. One model, taken from the business world, is that faculties or research institutes manage their own services in SOA domains with overall integration achieved through federated service registries and central monitoring and management of the network.
A major benefit of the SOVLE is that each tool can be removed and replaced by another in a simple manner. Integration with the SOVLE is governed by the tools interoperability model that we have developed at Leeds based on Shibboleth and SOCKET . This model, which differs significantly from current developments in the area, will be described in our paper.
One final point about the SOVLE is that the system architecture naturally leads to flexibility in the choice of user interface. Different interfaces will be used for different purposes. One graphical user interface that we are investigating for the SOVLE is a Project Wonderland Immersive 3D (I3D) world. We currently have a researcher developing an I3D interactive protein gallery for tutoring bioinformatics. We are also engineering some molecular modelling software so that it can act as a shared, collaborative application in our immersive world. Our paper will describe these developments together with the implementation of a “What is …?” learning design in I3D.