Loading ...
Sorry, an error occurred while loading the content.

Re: A content approach, and questions

Expand Messages
  • Tom Wason
    Prasanna-- This is an interesting thread. Here s my nickel s worth. Your desire to integrate resource management, workflow, tagging and taxonomies addresses a
    Message 1 of 9 , Jul 7, 2006
    • 0 Attachment

      This is an interesting thread. Here's my nickel's worth.

      Your desire to integrate resource management, workflow, tagging and
      taxonomies addresses a common situation. These are tightly bound.
      Where shall I begin?

      Resource Management: XML
      Arbitrarily, let me start with resource management. I think XML is a
      good thing to explore. It has several advantages. The elements and
      attributes can be used to guide various processes such as end product
      generation and context-based terminology searches for semi-automatic
      tagging. Take that last bit with a grain of salt. XML has the
      advantage of being ASCII text. This means it can be managed by a
      variety of tools, some as rudimentary (read "geeky") as CVS. It
      transports cleanly over the web (inter- or intra-net) securely though
      various well-known protocols (e.g., SOAP). XML content can be
      generated with open source tools such as OpenOffice (or its low cost
      better, StarOffice) that will also output in various forms (documents,
      HTML, PDF). This is essentially accomplished through the use of XSLT
      (XSL Translation) that can be customized. On a recent project to
      create a lot of reusable educational content in a US Air Force project
      we recommended the use of XML as the "source code." Outputs in a
      variety of formats were required. The system also had to conform to
      XML bindings of some of the educational standards. I have also used
      XML for the development of standards. In this case, the DocBook XML
      schema was used, with content management accomplished through CVS. As
      these were technical people working on technical standards, everyone
      was comfortable with this system. We used custom XSLTs for various
      outputs. (If any one wants to burn themselves out on this, I have some
      guides on my site that might be interesting,
      http://www.twason.com.drtomguides.html) . OpenDoc has replaced
      DocBook. The same idea only better tools.

      Creation resources will have some sort of workflow. At a minimal level
      it may be "tumble home." This is a popular programming method.
      However you are undertaking a program of resource modularization and
      reuse. This requires a workflow. I'll use educational resources as the
      model, as that is what you have referred to. Workflows are an
      excellent way to get at taxonomies and the process of tagging. I have
      found that people have a much better idea of what they do than what
      they need. Therefore, giving people a concrete task and asking them
      what they do, tracking what they do or watching what provides a lot of
      information about the taxonomic needs. It informs you about how the
      taxonomic terms will be entered and how they will be used. I have
      found this to be a very powerful method. One ends up discovering
      existing taxonomies, under other names. During the workflow the author
      may need to discover appropriate resources. That is the point of
      modular reuse. Some of the tags specifying the target needs may have
      been created such as audience type, learning objective, duration or
      size range and so forth. These values describe aspects of the context.
      These values should drop right into the search tool. The author only
      needs to add the information that is specific for that context.

      I am using "tags" in the broadest form, namely values in fields. A
      field may be a semantic structure. Let us not go down that path here.
      As you can see, tagging is bound up in the workflow. We have found
      that some tags are generated before the creation of the resource
      starts. For example, a brief description of the resource may be
      created. For an educational resource, the duration, target audience,
      learning objectives, MIME types, modularization and so forth may be
      generated. This is important information guiding the workflow, and
      serves as the basis for quality control at the completion of the
      resource generation. During the workflow additional tags may be filled
      such as author, modularization, associated assessment, assessment
      type, module MIMIE types, module topic, module title and so forth. The
      point is, if tags are created during the workflow, they have little
      impact on the workflow, are apt to be the most appropriate, and are
      created by the people who know the use.

      OK, here we are, working with what this CoP is ostensibly about:
      taxonomies. As you can see from the above, the taxonomies are the
      value sets for the fields. Yes, the entire set of fields and
      taxonomies comprise a taxonomy. We've broken down into sub-taxonomies.
      At this point, you have decided the format in which you will manage
      the resources. You have decided if the metadata will be embedded in
      the resource (yuck) or reside in a database. You have some idea of
      what the workflow is. You have deduced something about where in the
      workflow one might be able to capture and use information, the tags.
      Now you really get the end users involved more deeply. You want to ask
      of them:
      What do you need to know?
      What do you need to tell someone else?
      I have found that setting people work on a concrete task or set of
      tasks, in other words, using their workflow reveals a lot about what
      the need to know. You do not want to ask them what they need as a
      term, but what information they need. This information you then
      convert into a definition. At some point you will associate a name
      with that information, perhaps as it resides in a taxonomic structure.
      Let he users pick a name for each definition that they like. This is
      the local term. Some groups may create a structure. This is good.
      The users do not need to know about the core taxonomic structure. The
      tools should map the local perception of the taxonomy to the core
      structure. Note that the operative term is "map." This is not the
      same as a 1:1 thesaurus conversion. The mapping is semantic.

      I have put my comments on taxonomies last. You can see how they are
      buried in all of the previous sections, however. If you start with the
      taxonomy development, you will cover the other areas. The development
      of a taxonomy can be the nose of the camel under the tent. The
      development of a taxonomy can have deep impact on the workflow of an
      organization. It may even suggest a system architecture. As discussed
      by others in the forum, education is important. I have found that
      education has to be appropriate to the audience. Upper management
      will look to a cost/benefit analysis. It will have an eye to impacts
      across the enterprise. It will be concerned about impacts outside of
      the enterprise, such as sharing resources or processes among
      enterprises. The people who use the taxonomies need a different kind
      of education. They don't need to know much about how a taxonomy works
      or what it is. They need to know how it affects what they do. The
      taxonomists, on the other hand, want to understand the taxonomy as
      representing knowledge within the organization.

      Enough. I have prattled on far too long.

      As a closing note: Yes, a consult.


      Thomas D. Wason, Ph.D.
      +1 919.839.8187
      +1 919.602.6370 Cell
      1421 Park Drive
      Raleigh, North Carolina 27605 USA
    Your message has been successfully submitted and would be delivered to recipients shortly.