Dr Andy Sawyer, Simulacra Media
These notes were prepared for the Newcastle meeting of the AHRC-funded Semantic Web Think Tank project.
* The ‘Learning Object’ concept can be seen in the box loan schemes (used by museums since the 1880’s), providing objects which can be used for learning
* The conceptual foundations of digital Learning Objects can be found in Ted Nelson’s work in the 1960’s
* The term Learning Object first appeared in a paper by Wayne Hodgins in 1994
* Digicult’s ‘Learning Objects from Cultural and Scentific Heritage Resources’ (Thematic Issue 4, 2003) made explicit a key link: museum collections lend themselves to a learning object approach (see first point…)
There is no agreed definition of a Learning Object, but the ‘RAID’ acronym is a useful starting point – arguably Learning Objects should be:
— Reusable (able to be modified and used in many different learning situations)
— Accessible (able to be indexed and found as needed)
— Interoperable (operable across a wide variety of hardware, delivery environments and tools)
— Durable (continuing despite changes in versions of system software, players and plug-ins).
Most definitions assume small ‘chunks’ of learning which may be sequenced (hence IMS Content Packaging). An example of a Learning Object might be an image from a museum photographic collection, with some additional educational content and some metadata.
Q. Can we build a firm theoretical framework round something with no agreed definition?
3 METADATA & THE CHALLENGE TO CURATORIAL AUTHORITY
Metatada is a key to resource location. Arguably there are two approaches to metadata creation:
1) ‘neo-industrial conformity’ – resource heavy, top down, standards-facing, eg: IEEE LOM standard; Dublin Core; ARIADNE metadata specification.
2) ‘post-industrial’ – inexpensive, democratic, web 2.0, user facing: eg ‘folksonomies’ such as those used on delicious (http://del.icio.us/), flickr (http://www.flickr.com/) and youtube (http://youtube.com/) – browser bookmark sharing, image sharing and video sharing sites.
The rise of folksonomies and tagging could be seen as a challenge to:
— established metadata standards
— the hierarchic contexts from which they emerged
And curators may feel threatened – ‘its our stuff!’
Q. Is there a ‘third way’ to enable us to benefit from both classic or specialist taxonomies and volatile folksonomies, using for example OAI?
Q. Is metadata objective or subjective? If subjective, do we need to know who authored it?
Q. Is metadata produced once, or does it change as the data finds new uses in new places by new people?
In the 1990’s a range of standards were developed, mostly driven by the demands of the US military (establishment of Advanced Distributed Learning Initiative by US DoD in 1997, introduction of SCORM by Dexter Fletcher, working for the US military, in 1999). These standards are focused on managing learning objects in heavyweight software applications – Learning Management Systems, for example.
Q. Will standards and reference models from the above background will be useful given the rise of web apps, Google, the semantic web and social practice as evinced in web 2.0?
Q. Since standards only emerge after the event, how do we apply them to innovative types of learning object?
Q. Is it true that since education is “especially local, heterogeneous and contextual in ways that few other organized activities are … Local, situated knowledge … is not likely to be trumped even by the most well designed learning object or standardized practice for such a teaching process.” (Frieson & Cressman)
The definition in section 2 is pretty boring.
Can we have an ‘Agent Based’ Learning Object as proposed by Mohammed and Mohan, which:
— is aware of itself and its environment and should be able to respond to changes in its environment
— is capable of accepting input and exhibiting goal-oriented behaviour
— is able to recommend itself and act without the direct intervention of humans or other learning objects
— has control over its state and actions.
A good background which also covers standards is David Wiley’s, Learning Objects Literature Review (at http://opencontent.org/docs/lo-lit-review-draft.doc)
On standards and approaches to metadata, see N. Friesen and D. Cressman in A. Koohang, and K. Harman, Learning Objects: Theory, Praxis, Issues, and Trends. (Santa Rosa, California, 2007: Informing Science Press).
For some – by no means all – future possibilities, the work of P. Mohammed and P. Mohan, for example Agent Based Learning Objects on the