The 12th CETIS Enterprise SIG at Crewe Hall provided an excellent opportunity to kill a number of birds with one stone:

• A project review with the Reference Model technical QA consultant
• Catching up with Phosphorix's deployment of XCRI for harvesting course XML
• Presenting XCRI's work
• Making arrangements for the development phase with Phosphorix and London Met

Through Vashti's hard efforts, Enterprise SIGs have established a reputation for two things: practical presentations that give real insight; and novel venues. However, with Crewe Hall, it has to be said that Vashti excelled herself. Dating back to 1170 but rebuilt in 1663 and added to ever since, the Jacobean entrance hall with its surrounding gallery provided a truly inspirational setting. I wished I'd brought my camera (anyone get a shot of the entrance hall with the Christmas decorations?), but Warwick's phone gives a sense of its night-time Christmas wonder.

XCRI Project Review

Arriving early for the SIG created time for a pre-event project review session with Warwick Bailey of Icodeon, who is providing some technical QA on the Reference Model projects. Informed, insightful, thorough, fair and encouraging would probably be the best words to describe Warwick's approach to the review. Key plus points of XCRI's approach were identified as:

• Regular dissemination through the blog
• Emphasis on early release of practical deliverables for consultation
• Genuine commitment to dialogue with related projects
• Holistic approach to engaging key players and partners
• Evidence of real-world deployment

XCRI was also encouraged to continue its participation in the rigorous modelling work of COVARM and to ensure that as much experience as possible could be collected and shared from efforts to put XCRI's work into practice. XCRI was invited to see its plans to develop an xpath web service (see email conversation with Scott and Paul) in the context of a possible candidate for a service that had generic utility. However, the limitations of XCRI's current budget for pursuing this in depth were acknowledged.

Enterprise SIG Presentations

The event opened with a very clear presentation from Tony Hughes of West Cheshire College on the use of a standards-compliant service-oriented approach to integrating their institutional information systems. Careful mapping of business processes for student account creation, changing course, withdrawal, etc; and close collaboration with software vendors had been critical success factors. Although originally intended to carry more business logic, Microsoft's BizTalk server ended up providing a reliable store and forward service for IMS enterprise messages that were triggered by changes to data within the Student Record System. A custom COM object responded to these messages and performed the necessary actions on the college's network account directory. Plans were in place for other systems to join the service-oriented approach soon.

Questions afterwards confirmed that West Cheshire's plans for future development in the area of tracking learner engagement were a shared priority for members of the Enterprise SIG and that Business Intelligence and Customer Relationship Management in the e-Admin domain should be funding priorities as standards for collating such information from disparate systems were immature at best and for the most part non-existent.

The next presentation was from Selwyn Lloyd of Phosphorix who described how work developed for the JISC-funded SHELL project had been developed to provide a platform for two regional pilots: the Liverpool Learning Matrix (LMX) and the East of England Lifelong Learning Support project (EELLS). The reliable messaging ioAgent/ioNode platform developed in SHELL had been used to provide the backbone for harvesting course information from partner institutions to create an aggregated catalogue of course offerings. Exchange of course information between nodes had been done with XCRI's draft XML.

Selwyn provided live demonstrations and Anthony Beal of the Liverpool LMX talked about the project's impact on learners. An E-Portfolio/CV builder tool had been developed as part of the EEELS project, which had provided an opportunity to reflect on the links between learner achievements and the awards and entry requirements of learning offerings. It was clear that valuable experience had been gained in using an early draft of XCRI's XML in this context, and XCRI's Programme Manager confirmed that some of the remaining budget could be re-directed towards documenting this experience and using it to support development of XCRI demonstration web services.

After lunch, Antony Corfield gave an update on the University of the Highlands and Islands' BEWT project (Bodington integrated with Enterprise Webservices Toolkit). Antony's presentation provided some useful tips on operationalising the CETIS IMS Enterprise Webservices Toolkit, including platform versions required to run the demonstration software. Questions inspired a useful exchange on Java technologies for persistence - BEWT was working with Enterprise Java Beans, and participants recounted favourable experiences with Hibernate. For java implementations of XCRI's webservice that do not use an XML database, a suitable persistence technology that is neutral of the underlying database implementation will be important.

XCRI then provided an update of its current progress and plans. It focused in particular on the emerging information model; the major curriculum fragments that might be re-purposed: spec, offering, recognition(Of Completion), benchmarks, learning outcomes, and org; and a flexible supertype for relating curriculum fragments, comprising selection and collection, which could be realized as associated subSpecs, corequisiteSpecs, excludedSpecs, awardingOrgs, accreditingOrgs, teachingOrgs, etc. There was general support for the conceptualization and SIG members were keen to see a common Information Model emerge from collaboration between Reference Models (XCRI, COVARM) and Regional Pilots (Pathways4Progression, SUNIWE). SIG members were thanked for their feedback on the v0 XML schema draft and interest in the v1 release was clearly high.

The SIG concluded with a presentation from COVARM that emphasised the value of a model-driven approach that went all the way from interviews with administrators involved in Course Validation through to the specification of webservice interfaces for components that would support the validation process. A simple costing reminded those present of the e-Admin value proposition: save time to re-invest elsewhere in the business of learning.

XCRI Development Plans

XCRI is funded until the end of March, 2006, and remains committed to delivering a working demonstration of a curriculum webservice within that timeframe. Re-visiting the information model to get alignment between other JISC projects has inevitably introduced delays, but XCRI still hopes to release the v1 XML schema by Christmas 2005 so that work can begin on a webservice demonstrator first thing in the New Year. XCRI's Programme Manager has approved a plan to redirect remaining budget to engage the services of Phosphorix and Paul Walk of London Metropolitan University to create, test, document and publish the open source demonstrator. Crewe provided the first opportunity for Paul to demonstrate an emerging java servlet prototype that retrieved curriculum XML from a Berkley XML database.

Given the timeframe and budget, scope of the build will be limited to retrieval rather than updates. The issue of updates and authoring technology must be tackled in any follow-on work.