Friday, 11 February 2005

Metadata framework

I attended the Melbourne !dea metadata workshop chaired by Jon Mason yesterday. The workshop is covered by Derek's Blog. See here and here.

The first part of the workshop, we heard from leaders of current DEST funded projects on their metadata implementation issues and experience. It was eye-opening.

The second part of the workshop, we brainstormed, trying to start the process of developing a "generation-2" metadata framework in response to the ISO SC36 initiatives. The e-learning Australian Standards Committee (IT19-1) will be drafting our Australian response in the next few months.

Being a framework, I believe we should not get bolted down by the details. We should examine the overall metadata requirements of all educational institutions. We should then "dice and slice" the requirements and identify those which should be deal with by the "education community" and those which may overlap with other communities of practice. I further believe that we should concentrate our effort on those educational specific requirements and collaborate with others on the common parts.

Several different ways of dice and slice were suggested. At this point, I still prefer my own version which may change when I think it through again.

Here are my board grouping of the functional requirements of metadata in order to support the metadata needs of any educational institutions.

  • Support Resource Discovery

  • Intellectual Property & Digital Right Management

  • Learning Objectives & Competency

  • Context Descriptions

  • Trust-related Descriptions

  • People1

  • Preservation requirements

  • Resource dependence

  • Annotation/management of metadata

  • Linkages and aggregations2

  • 1By people, I meant both the intended audience and the role of the creator when the resource is generated. Let consider the role of the creator when digital data is generated. What I am trying to say here is that we should not narrowly define educational resources to those generated by institutes and "teachers". Students generated data, during their learning process, may have re-use value that some sort of metadata attachment may be appropriate (especially when the cost of generating and storing these metadata are near zero).

    1. as a record of student's work for subsequent verification (may be for qualification or for later skill matching)

    2. for researchers to analysis the learning process

    3. for researchers to analysis the related content. For example, for item analysis in item bank in formative evaluation.

    4. The output of Master/PhD level (strictly speaking is student's work) is of high value to the global "knowledge pool"

    2may somehow expose citations/bibliography within the resource in order to trace development history of the subject domain (e.g.)

    There are significant overlapping of elements which may serve more than one functions in the above scheme. In order to ensure orthogonality of elements, some significant decisions and discussions need to take place.

    One of the thing I like to do in the near future is to apply my meta-meta-meta data model and see what I can get out of this in the education context. Please keep watching this space.

    No comments: