Doug Macleod – The Future of Learning
- Better experience for learners – when I see ‘Return on Investment’ I want to say that providing better learning is enough
- Create lots of high quality engaging materials – doesn’t matter if they are interoperable, doesn’t matter if the keywords don’t work
- EduSource – create a testbed of linked repositories, create community
- Facts – dates, budget
- EduSource – will be based on standards, bilingual, accessible (with Jutta, TILE)
- Sharing and disseminating all findings
- EduSource – won’t provide production level service, won’t provide content, not restructed to broadband networks
- Partner list
- Layers for learning – repository layer / communications layer / communities of use / …
- Repository in a box… - make it simple and accessible
Gary Popovich - Meeting challenged in the metadata world
- or: getting and keeping our own backyards in order
- ie., we need common understanding, common language, leadership & innovation
- Alberta initiatives: LearnAlberta.Ca, CAREO
- CMEC – Pan-Canadian repositories initiative
o Making sense of the metadata world
o Decisions around standards, reaching agreement
o Implementation of these standards across stakeholder sectors
- Learning objects – defining them (issues include contentualization, levels of granularity, meta-tagging)
- Designing a learning objects:
o Instructional design, functional design, pedagogical quality, content, quality (400 page manual)
- Protecting intellectual property – coping with the high cost of developing, licensing, acquiring digital content
- Revenue models: there isn’t any
- Content quality, selection, exchange – need for standards, and who sets the standards
- Alignment of effort and direction at the provincial level
o Convergence of CAREO, LearnAlberta
o Technology standards
o Alignment of learning objects and metadata standards
o Co-development and co-sharing of content repositories (East-West project) – it was the activity that brought us to standards
o Implementation of a learning and technical policy framework
o Development of ‘tool-kits’
o Research and best practices – delivery, access, etc.
o Volume licensing and acquisition of learning objects (Comment: ick, no, don’t do this)
- CMEC portal repository initiative – one-stop gateway for educators in Canada, staged development and implementation – the common portal links them to other portals – will put out an RFP for a project mgmt group to build the portal – pan-Canadian acquisitions and licensing (Comment: Why would you do this?)
- Agreements on technical standards, shared RFPs, eg. Alberta and Ontario
- Stages 1 and 2 – October, 2003 – Stage 3 – October 04
Chuck Hmilton – The Aggregation Game (from a ‘confidential’ slide)
- We need to start building on the outer rings – building the part where the humans come in
- Someone said – “We’re all going to be librarians and we’re all going to be authors”
- All that content out there – who’s going to put all that into packages – are we going to put our 80 gig drives online? There’s a missing piece out there
- POOL – like grain pools – but we haven’t gotten into personalization, flow, presentation
- Content vs Context (content declines, context increases – 85 % of content will have to be context sensitive within five years – JD Levy)
- Levy: Get out of prisonware
- Levy: all learning today is like a light build where everyone gets illuminated equally – moving toward laser light learning – directed
- I come to you as a reporter – we field requests – and these are the kinds of systems & solutions that are being asked for – but these are difficult questions
- Eg. 60,000 users are the base for a trial – LCMS movement is there, but the infrastructure is shakey
- Let’s move IMS from the background to the foreground – identify three exemplary projects – “Make people eat your oreo cookies” – demonstrate how IMS is tied to the business models & watch the flow
- This year, IBM will make a giant stand on where it’s going regarding learning
David Porter – Repositories R Us – A case for user-centric repositories
- One of the most powerful strategies in education is modeling
- If the usual suspects truly believe that repositories are the way to go, then I challenge us to be repositories
- But to do that you need some really cool tools to make that happen
- When Napster came on the scene, we needed to wake up and take notice – 50,000,000 users can’t be wrong
- Yet I’m still looking for a learnster
- The kids that Napster represents are the consumers of the future and the learners of the present:
o Try before you uy
o Buy by the piece
o Mix and match
o Etc.
- Many simple tools with simple metadata and intuititive interfaces that allow you to build a personal collection that you can share
- Convivial tools – it’s immediately obvious what they do and they’re useful to you now – eg., various media desktop
- Some examples in academia:
o Splash
o Recombo’s converter tool
o Indiana – possibility network
- If I’m a busy academic I want simply tools I call up during my workflow
- What will the environment look like:
o Splash – the locus of repositories should be personal first
§ Then make sure the metadata is accessible
§ Give me a simple desktop app that is always accessible
§ Layers of metadata – if I only want 3 or 4 items, that’s fine
o Recombo Converter
§ Converts legacy content (eg MS PowerPoint) into new formats
o The Possibility Network
§ I’m not building a portal, the portal’s on my maching
§ It goes out and searches and brings that back to me
§ Based on a web services notion that there will be a network of providers, but the portal will live on my machine, and I will shape its operations and activity
§ Personal learning strategy designers – not a set of tools that are centrally located on a server
- We are not paying enough attention to the end users and we need to focus again on what the web is all about
Questions
To David Porter: Students will like it, but I have to sell to faculty. Porter – these tools will also apply to faculty – they can talk to them in a language they understand. Tools should be able to be skinned by you or by faculty. Tools should be something you do yourself rather than something that’s done to you. Rejoinder: would not sell, because the instructional design will be bypassed. Response: not my message. This can be built in.
To David Porter: Right on. The notion of a Napster type environment is tremendously appealing. Is Splash real? Answer: yes it’s real. The developers are here. It is not perfect but it’s a step in the right direction. And it has quality assurance features. Rejoinder: How do you reconcile that with copyright? Answser: I don’t think it’s a problem. We want a tool with the qualities of Napster, but not divorced from a DRM scheme. Open source, public domain and digitally managed rights can exist on the same network. Rejoinder: Is there a central location where this info is being published. Response: Eduspecs.ca along with edusource.ca and cognisource.ca
Point about faculty: the term learning objects is deceptive because it implies atemporality – for faculty, the interesting thing is moving development form a big umbrella to a syntax for assembly. Tools need to reintroduce the notion of iterability to learning objects. Response: we are working on these types of concepts. I think the notion of the stand-alone decontextualized resource is a real turn-off. We will be asking, what’s an object. Is object reuse a myth? Another response, from Doug MacLeod: there is a variety of tools that have been developed. Industry Canada is helping through its pilot projects to get those tools out into communities to be tested.
Question to David: one of the things we want to do is learn how to learn, and become our own instructional designer.. Response: yes, but I’m a guest here and I didn’t want to offent. If you need a manual then there’s something wrong with the tool. Mix. Kids already know how to do some of this stuff. We need to simplify the models and widen the circle of participation.
Question: to Chuck – how do you reconcile the societal view of learning with the personal one (ie., David vs. Gary)? Answer – it’s a continuum – there will be people working individually, and people working in a group. What I we could say to faculty, ‘we have a way of metatagging what you know so we could share it’. Then we could work across any community. I don’t think we have to bring them together.
Question: the idea of communities is essential. People learn in groups. But they also learn personally. But how do we tie them. Faculty – couldn’t they use a tool like this? What we really need to do is break down the walls. How do we do this as a society, because it really has to do with how we reshape ourselves. I want like to draw the American research communities in over this notion of infrastructure – NSF notion of infrastructure, but it’s all about research, there’s no learning built in. Chuck: we are all working in the most fragmented community in the work. We maybe need leadership to take the stage.
Question: David’s version vs Gary’s vision to standardize across the 10 provinces – how do we standardize across all those areas we can’t agree on. Gary: Isn’t time up? I don’t think we are going to standardize, but we are going to come to agreements about what we think are effective practices. We’re not going to take a cookie cutter and say it’s the same for everyone.
Norm Friesen – Participation in Specification / Standards Activities
Participating memberships in international standards bodies
LearnAlberta.ca
- audience specific in its orientation
- learning objects will support Alberta programs of study (specifically, curriculum outcomes)
- Designed to support lifelong learning in Alberta
Content Packaging
- intended to create standardized set of structures to be used to exchange content
- Alberta Learning uses the organizational parts of the spec
- (Screen shows, eg., National Geographic stuff)
- Content package is downloaded as a single unit, user browses through several organizational levels to get to a specific resource
EduSource – packages and related work
- Specifications – OAI metadata harvesting
- But also – learning communication layer – gateway architecture
- EduSource UML diagram
- ECL – eduSource Communication Protocol
- CanCore
o Application profile – customization of a standard to meet the needs of a particular group
o Subset of LOM activities
o Guidelines document, best practice recommendations (175 pages)
o Effective metadata requires semantic specification and consensus
o Incorporate best practices from library and heritage communities
o CanCore the only body doing this across projects in different settings
- We need to take a broad overview of what’s going on in application profiles – some elements are very common, others are used inconsistently (very diverse pattern of use – great diagram)
Peter Hope – International Conformance Program in Canada
- Interoperability is still a big issue with us because we’re a small market – how do we leverag for an international audience
- Do things really work – that’s what conformace is all about
Why conformance?
- Purchasers want to compare an apple with an apple, not apples and organges
- Reality today: multiple LMSs, multiple content platforms – today if you’re lucky they’ll join, but more often than not, they won’t
- Vendors don’t want to always be going to conform to some government’s spec – why not go to a standards body where I can pay to have my content tested & I can put a sticker on my box – this is what electrical devices do
- This lowers the entry barrier
- SCORM through ADL has a self-test suite
Canada
- Our application profile – our interpretation of how to make IMS work for us – now will implement the IEEE profile
- But if my stuff is CanCore conformant and the purchaser requires that it be SCORM conformant, what do I do?
- In Canada, we also have issues with other languages, especially in a NATO context
-
Dick Hall
Remit
3 year program – to establish foreign markets for British universities
- have developed a platform
- The bottom line is key: quality – the only way we make this work is through our reputation
Why a new platform?
- Limitations of existing virtual campus products
- Limitations of corporate products
- Designed for adult HE learners – more than page turners
- Support for course development teams
- Open system architecture
- Scalability
Platform design objectives
- must be modular – for improvements, adaptation, sharing
- must be standards-based – the whole focus is the ability of the student to learn using our platform
- pedagogically-prioritised functionality
- support for team-based development – if you want to produce high quality, you need a team and workflow
- scalability
- connectivity with university MIS
- world class
Elearning standards
- IMS
- SCORM 1.2
- Open Knowledge Initiative (OKI)
Accessibility
- should be “leading edge” in promoting accessibility – blindness, deafness, etc
- following best practice guidelines
Why are standards important tol you?
- platform (hard) or content (easier)
- compatibility, cost, ease of implementation, internal policy, choice
- what standards are important to your RFP?
Things to look for in a supplier:
- Do they claim to be standards compliant – do they understand them in your context (it’s more than saying the right words)
The Holy Grail: a widely accepted international elearning conformance programme with an accredited independent testing regime
Tim Magner – Schools Interoperability Framework
SIF – nonprofi organization – more than 100 K12 tech companies
SIF zone – each application has an agent that takes proprietary data set from application and translates them to other apps through a message router – platform independent, vendor neutral, intended tpo leverage existing work
Messgae types – add, change, delete, request/response, asynchronous messaging, etc
SIF adoption and development
- portability vs interoperability – both data structure and messaging structure – ie., must support both data and delivery of the data
- service and support infrastructure – agent development kits, agent development services, implementation guides, seminars, workshops, product lists
Compliance
- 3rd party evaluation
- allows confidence, validation
- we test validation conformance
- should be available next week
- trademark and license agreement
- ‘test harness’ – testing takes place over the web, tests all functional messaging, all traffic logged
- documentation sent to The Open Group for certification application
- Legal agreement between vendor and the Open Group to ensure continued compliance
- Then the vendor is granted a trademark license
SIF and IMS
- profiling applications: avoid redundance, consistent set of standards, leverages existibgf work, provides real work feedback
Government Encouragement and Adoption
- “a little like a 7th grade dance” when it comes to government endorsement
- does SIF meet the needs of various governments – other governments with similar needs
Specification Evolution
- driven by customer requirements
- K12-verticals – applications (eg library, etc)
- Great data model graph
- Vertical data pyramid
- Seamless reporting – district, state, federal
- Reduce burden on schools and districts
- Etc.
SIF is a technology solution in support of a management solution – enterprise wide data management is vitally important
Active Content
Peter Hope
Need – training – “let the person play” – on an expensive airplane or an inexpensive system
Gilbert Paquette
Main innovations:
- departs from the dominant authoring paradigm, getting ino events, etc where people interact
- actor patterns, multiple models
- visualization, annotation, adaptive advice capabilities
- actual workflow inferface
Learning ecosystem
- multi-actor workflow (diagram)
- this is the kind of system we want in an open university
- the same sort of graph can represent a project in a PBL environment
- so we need new instructional design approaches
New approaches
- knowledge management – knowledge intensive learning
- nano-seconds development
Who or what should be active?
- content objects are important, but how do we use them
- separate content and use
- have we talked enough about activities?
Active content – a typology:
- Interactive objects
- Adaptive objects
- Adaptive agents
- Active environments
- Organic learning ecosystems (where we want to go)
EML / IMS-LD concept
- uses a play metaphor
- alternative roles and activities, but the roles are not fixed
Challenges
- move from one end of the typology to the other
- extend LOM, SCORM specifications
- ease the designer tasks
- provide interactive multi-actor workflow interface
Paul Stacey – People to People, not just People to Content
Bringing people to content – hence learning object repositories, metadata – most of this focuses on autonomous self study
BORING – static content
Greatest value of e-learning is people-to-people eg peer-to-peer, we want content that can change, content that is current, content that can evolve
Less emphasis on prepackaged content to creating a framework
Let’s get more people involved
Most of our discussions have focused on LMSs…. A lot of it is COTS and trying to hook all this into enterprise systems – and we get lectures, course notes and resources –
This is necessary but not sufficient
We want IM, weblog, webcasts, RSS threads, etc – the content produced by these activities will be collaborative, cover live events, game based
Involved in…
- Learning Times – online learning communities – blogs, boards, threads, etc
- SFU – enabling peer-to-peer group based activities
- Poetry slam voice board (yeah!)
- Active and alive exemplars – cognisource – enabling as much pull and push as we can
- A lot of e-learning today is push-down from the creators – the market pull, though, is less for learning objects than for interactions & communities
Does learning require content
Peter Sloep – active content and IMS learning design
A Play…
- consists of a number of acts
- acts consist of role parts – many different role parts in an act
- each role consists of (a) role (b) activity
- each activity is connected to a role, and takes place in an environment
- in the environment, there are learning objects and learning services
-
Forms of active content:
- Static
- Dynamic – decisions
- Mechanisms that define decisions are properties – properties are variables, which may be defined at will. Properties result from student interaction with the system
Structured Content
- properties are set from within learning object
- interaction requires access to the interior of learning objects
- use XML – structuring content – this allows you to insert other namespaces such as QTI, MathML, XHTML, Doc-book – media neutrality
Tom Carey
We just had a wine tasting session :) but of course in any good wine tasting, you can’t swallow much
- Peter Hope – we can get from this Rose where we are now to a robust Red
- Gilbert – presented a complex wine with many elements of taste, long aging process – what can be separated? Can we separate content from use? Where is the nouveau beaujole
- Paul Stacey – power to the people – represented BC pear cider – the working person’s drink – does learning require prepackaged content? Or, I say, can someone prepackage routes to that learning? And what about the management of user-created content
- Peter Sloep – gave us a course in wine-making – I thought it was a very nice and spritely white wine – it’s very nice, but I’m looking for a red – eg. I want actions that take place on the non-completion of activities – requires that concepts like time be given first class citizenship – again, can content and use be separated?
Peter S. – simulations are going to run in their own environment – we want the data fed back into the LMS to influence the flow – the simulations can be developed in relative isolation.
Tom – is active content a more active management system or a more active content – Peter S. it’s where the properties are accessible in real time, so activities can talk to each other.
Peter Hope – but where do you declare the properties. External to the object, or internal to the object. Dumb objects or smart objects? Peter S. That would be an issue. The properties might not be part of the package, they might be part of the student’s dossier.
Question – reference to “the autism of knowledge management” – what has come out of this whole day essentially is the notion that there is a feedback loop we should be talking about between designers and users and a community of objects… the consumption of objects is not just an in-and-out thing – it’s an iterative process – you change the object by interacting with it – the tech of instructional design has leapt forward irrespective of the tech of learning objects
Gilbert – content and context of use – if we just do contextless use, then we won’t get drunk, we’ll just get sad and fat. You could have context of use that is very directed – you tell the learner what to do – but sometimes I am wondering whether learning design doesn’t look like programmed instruction – or you put them into a different environment, it’s still the same object, but in a different situation. Or perhaps use the same object in a teacher training class, where they analyze the object from a pedagogical point of view. Learning requires content, and sometimes the learner produces the content. If nobody has an agenda, well, we know what happens in forums when there is no agenda. Some planning si required. But I am against fixed planning. Add another option – emergent flows. Just like what happens in a project. We start with a plan but then we change it.
Peter S. – I agree, that’s where we’re going. No reason why we can’t have intelligent agents that operate in a community of leaners, that create designs and feed back designs to the learner. Emergent systems. Self-organizing community. A book is valuable. So is a community of practice.
Paul Stacey: how many of you prefer to drink alone, and how many prefer to drink in a social context? I would prefer to be thought of as a basic house red. Look at the time taken creating content – about 70 percent taken creating prepackaged content, only 30 percent on the rest. Turn this around. Have partially created content, have students finish it. Why not let students that come see the footsteps of students that were there before them.
Question: we are painting ourselves into two corners: 1. the ultimate playpen vs. 2. the need for having outcomes that are predetermined. We need to serve both of those possibilities.
Gilbert: yeah but you can reach that goal without telling telling telling people…