XplanaZine

Session 2
Friday 10:45 - noon
Session — Best Practices - eduSource Canada
Presenters — Mike Mattson, University of Calgary
- background: pan-Canadian research project, with the goal of creating a testbed of interoperable learning objects across Canada; also provide a forum for the development of tools; standards for infrastructure
- project background: began July 1, 2002 - March 21, 2004; second phase of funding as part of CANARIE e-learning program (50% funding model); total budget of $8.5 million CAD
- founding project partners: newMIC, Athabasca U, Netera Calgary, U of Waterloo, Technologies Cogigraph, Tele Education NB - instead of funding each individually, with individual research areas, grantor wanted them to work together

- academic partners: lots! Good buy-in from Canadian universities
- industry partners: is there an economic incentive for this?
- a case for collaboration: create a critical mass of national expertise,share research to avoid redundancies to avoid repetition of mistakes,leverage existing and planned investments (bring down the silos)
- vision: founding partners were ready to go with tech, ideas - but needed a vision; worked with Stephen Downes on the vision committee - “Enable Don’t Require”: implement international standards and specs, vision, tech, sustainability; provide open source for the community
- vision to reality: evolving specs; open source issues - uPortal has cost $300,000 to implement so far at their University, have to take into account the whole cost: shared (community) source vs. open source makes more sense, but you need a model of sustainability; many right points of view; 50% funding model
- work packages: repository content development facilitation (education is provincially administered, not federally); DRM - huge issue, Stephen Downes working on it (looking at the iTunes model); testing and evalution; metadata development - best practices and guidelines to implement IMS; software development; hardware integration; business and management models - looking
at what sort of open source models can be adopted; community building - run out of Waterloo, is a key part - “build it and they will come?” no, “build it and you have to go out and sell it” to key stakeholders, have to make the case; project management and co-ordination
- megadata: professional indexers - controlled vocabularies, learning outcomes (for example, tools to access the learning outcomes for the k-12 system and attaching it to learning objects); educators - simple needs; CanCore application profile - needs refinement
- standards: building to standards and specs to achieve interoperability - “lies, damn lies, and specifications”; as implementers of the standards, need a way to go back to IMS for clarification of specs, or a way for rapid prototyping of specs
- software development: metadata repository, resource management services, digitized resource repository, DRM, e-learning middleware services, communication kernel, tied into other LMS systems
- eduSource Communication Language (ECL): connector based on IMS DRI spec, with SOAP messaging, OAI, Z39.50; gateway for translations - design and build once; each repository has to create a single plug-in for ECL; doesn’t need to be open source, needs to be extensible
- why not use a search engine, rather than a repository? Richer, more useful metadata
- eduSource infrastructure supports 3 types of users: federated search; individuals via peer-to-peer (selective publishing); harvested - intense debate between the three, only made sense once it got scratched out and people saw it could work together
- database conundrum: xQuery - few implementations in open source databases, relational database support; native XML vs. relational
- database solutions: xQuery, creation of search channels by interest groups (math repositories, science repositories, etc.) to narrow searches
- eduSource repository: available services will be published for users to access; administrative issues to be addressed (policing content, how long will it be available, etc.) - epository-in-a-box
- eduSource challenges: scope, distance, conflict resolution, funding model,managing expectations (prototype to production quality)
- campus initiatives: CAREO (educational object repository at U Alberta) - display, search, browse objects by discipline, access metadata, subscribe to objects, view objects in its own window, view object history, metadata entry and media upload, discuss/review objects and peer review
- Stephen Downes challenged IMS: learning object repositories need to be as successful delivering info as RSS - now, RSS feeds are available for new objects, my objects, etc. in CAREO - one of the single biggest accomplishments of the whole project
- guest speaker: Alan Levine from Maricopa college, to talk about their RSS and learning objects initiative - they don’t consider their project a repository, more a collection of descriptions - a “virtual warehouse”; no metadata, no LMS; the metaphor for a learning object is a package: each
package has a packing slip with info, can upload media, can add as many links as they want; meet a need for improvement in how info gets used, how info gets aggregated, how to include small packages of info in other branded sites using RSS via javascript; using weblogs to describe the context of a
learning object (David Wiley), to put a personal take on how an object is used - so, each object in the MLX has a unique identifier (RDF) that gets used to describe the history of an object used between blogs via trackback to connect the object with the way they’re being used
- building a bridge to embed CAREO learning objects within Blackboard, within portals - quicker and easier with RSS
- reusing the repository - using Web Objects for creating departmental and institutional site themes (for example, http://www.sciq.ca), a challenge for designing across browsers
- links:

• eduSource (http://www.edusource.ca)
• CAREO (http://careo.ucalgary.ca)
• development blog (http://commons.ucalgary.ca/weblogs/dnorman)