Subject: Re: Learning asset and learning object
From: Albert Ip (albert@dls.au.com)
Date: Fri 03 Mar 2000 - 04:32:49 MET
From: "Albert Ip" <albert@dls.au.com> Subject: Re: Learning asset and learning object Date: Fri, 3 Mar 2000 14:32:49 +1100
List address to send message to everyone: ifets-discuss@LISTSERV.READADP.COM
Details of current discussion: http://ifets.ieee.org/discussions/discuss.html
---------------------------------------------------------------------------
On Wed, 1 Mar 2000 15:28:42 -0800 Clark Quinn wrote:
>There are two types of tagging. XML tagging can be on the 'outside'
>of the object, and then it's just other tagging. You can also use
>XML within the object, but that's essentially opaque to the system
>(unless your system includes more than any LTSC requirements). It's
>certainly likely to be opaque to any arbitrary IEEE1484-compatible
>system
You probably have read "Embedded tagging is harmful" (?? I can't remember
the title exactly, but it was on a special issue on XML in w3j) when you say
XML tagging "outside" the object. However, I don't think the current
standard on XML are embedded tagging. Again, we may have misunderstood each
other in the use of terminology.
Metadata as applied to resources can be embedded and detached. I don't
refer to metadata as tagging in XML sense. XML documents have tags. XML
tagging refers to the act of applying tags to the whole document (making the
document an XML document).
When and if educational content are marked up (tags applied) as XML
document, and when and if software modules which know the tags are
available, a sub-unit of the document may be used by the technical system.
However, personally, I don't have sufficient faith in IEEE-1484 compatible
system being able to handle the diverse contents currently being used in
education and training. A straight-jacket approach is not an ideal approach
for me.
>>"It requires metadata tagging to be a fully realized object." - this is
>>neither the necessary condition, nor the "sufficient condition". I can
>>design a system to use a resource without requiring the resource to be
>>tagged (both metadata tagging or XML tagging). Equally, I can tag any
>>object and it does not automatically make the object a fully realized
>>object.
>
>I'm talking about the standards. What you say is true, in some
>sense. I'm talking about ways to make interoperable systems (which I
>perceive as one of the benefits).
My view of interoperable systems may be quite different. I like to use the
plain old telephone system as an hugely successful interoperable system
example. We have different made of telephones operating throught the world.
These telephones can have quite different features - some external to the
"telephone system" (like recording and leave message) and some intrinsic
(like conference call). They all work and work interoperably.
Unfortunately, learning system is not a simple system (in a way the
telephone system is simple because its main function to transfer voice from
one end to another). By defining a set of "functions" and requiring all
content to be marked up according to a standard to achieve interoperability
is a least common demoninator approach.
>>Metadata tagging supports resource discovery. How rich you want to go so
>>that the tagging also describes things like activation parameters,
variables
>>settings, typical use... ? How complicated the software will be to take
>>advantage of all these "rich tagging"?
>
>I think these are two ends of the continuum (as Crispin Weston has
>also indicated); you can have simple objects with rich tagging, or
>smart objects which have rich standards for information passing to
>activate them to the context. I think simple objects with rich
>tagging is a more vendor-neutral approach, and that's the approach I
>believe the LTSC is taking. I was merely elucidating that framework
>within which to make my case, not necessarily defending it.
Again, I think we have misunderstanding of terms here. Metadata tagging is
what I called type 2 data (sorry, I think I need to point you to my paper in
order to have a common understanding of our view points. How shameless!
http://www.dls.au.com/metadata/DataModel.html). Such metadata may be for
machine understanding of the "whole" resource. Smaller grain-size re-use,
such as taking a component out of the resource requires that the resource is
decomposible. One such mechanism for possible decomposition is XML markup
of resource (some people refers to XML markup as tagging too, I think that's
the misunderstanding!) with appropriate DOM (document object model).
LTSC approach (correct me if I am wrong and I really hope I am wrong) for
vendor-neutral interoperability is by defining LOM metadata standards so
that LO can be discovered and hopefully by using the metadata, there is a
level of machine understanding of the feature of the LO to make use of it.
It is re-use on the grain level of complete LO. However, if such LO is
content and such content has specific software interactivity requirement, a
vendor-neutral system would not be able to use the LO. So, work in IMS, for
example the Question and Testing group, assumes and requires the content to
have XML markups. Unfortunately, the tags need to be standardized as well
in order to have interoperability. However, this time, we may have re-use
in a sub-grain level.
I was struggling with this issue couple of years ago and have since proposed
the "virtual apparatus framework" which may overcome this issue. (What I am
doing today! so shamelessly advertising my previous works!)
>>If your version of metadata tagging supports "some annotation about
author,
>>time period, style" and "audio overlay" that accompany a chart to explain
>>it's context, I think your metadata tags are too rich to be useful. In
fact
>>such content of the tag is better separately stored as a resource. In
this
>>case this separated stored resource is an "learning asset". If I can find
>>this learning asset, I can know ONE example of using the resource referred
>>to by this learnign asset. Hence the learning asset becomes an enabling
>>device for using resouces that are not originally created for education
>>consumption. (This is a long phrase, can I use "NEF resource"?) But
still,
>>tagging (or metadata tagging) does not change nature of the NEF resource.
>>There may be other education use of the same NEF resource. I think that's
>>OK and is prefectly welcome!
>
>I think we're coming up here against a definitely difficult issue:
>what separates a knowledge/information object from a learning object.
>The more I think about it, I think there's not a clear line of
>definition. I've made the initial claim that it must have an
>implicit objective (in the rich sense of learning objective, see
>below) to be a learning object. I may be wrong.
Yes, I am struggling with the same issue for a long time and never have a
satisfactory answer. I am glad that this discussion have highlighted the
issue and make me clear about one or two points. However, I think we still
have a long way to go.
>> >If there's an external context provided, I'll agree. But I don't think
>>that serving a string of independent knowledge objects will necessarily
>>yield a learning experience. Currently. That is, a system that can
create
>>a learning experience on the fly out of tagged knowledge objects just
>>doesn't exist yet, nor can it unless some more breakthroughs are
>>accomplished in AI and cognitive science.
>>
>>
>>People do not learn in vacuum.
>
>I think we're agreeing furiously.
>
>> >OK. Given the above, I'd argue that content, functionality, and 'look
and
>>feel' are properties of objects, period. Knowledge or learning. The need
>>for a learning objective (implicit is acceptable) is what's needed to make
>>it a learning object.
>>
>>
>>Is it that simple? Just learning objective?
>
>No, not just that simple. A learning objective implies getting a
>particular audience to achieve a particular level of performance in a
>particular context. It's the learning 'wrapping' around some bit of
>data or information that makes it part of a learning experience. The
>contextualization you're talking about above.
Yes, my point was learning objective is ONE of the many requirements in
order for something to be qualified to be a LO. The concept of "learning
'wrapping'" is fine as long as we know that is that wrapping.
Further thinking along this line? I appreciate working together to achieve
a mutual understanding.
>Thanks very much for your detailed interactions, -- Clark
me too. I have enjoyed every bit of the interaction. thanks for creating
this great opportunity.
Albert
---------------------------------------------------------
Forum website: http://ifets.ieee.org/
Forum's contact person: kinshuk@massey.ac.nz
Info on Join/Leave List: http://ifets.ieee.org/maillist.html
---------------------------------------------------------
This archive was generated by hypermail 2a24 : Tue 07 Mar 2000 - 04:03:35 MET