Boston KMForum

To Tag or Not to Tag!
Should we be structuring our Knowledge Assets?
Is Free-text Search Enough?
Attendee Discussion, Moderated by Lynda Moulton

  • Can you fully exploit Knowledge Assets in the form of documents without categorizing them?
  • How contextual content is best understood and presented?
  • Is automatic categorizing working for some (or all) of your content?
  • Does taxonomy or navigated search have a place in your organization? Is it available and used?
  • When is key-word, free-text or Google style searching good enough?
  • Are you clear about which approach is needed and for what circumstances?

This debate has endured for years but the technological ground keeps shifting. Business managers and audiences are becoming more sophisticated about search, where and how they need to use it. We want to hear how you understand the issues and who is controlling the discussion in your organization: IT, library, vendors, or users.

See PowerPoint with meeting commentary on the points listed above.

Boston KM Forum thanks Jarg Corporation for hosting Thursday meetings

From the field

Phil Murray, KMConnection

Wish I could be at the March 16 meeting, because I have a lot to say on the topic…So I’ll throw in my 2 cents by email:

Your first question—“Can you fully exploit Knowledge Assets in the form of documents without categorizing them?”—is deeply flawed.

Let me quote from an article I’m working on:

“The fundamental premise of applying semantic metadata to enterprise assets — whether expressed as a simple vocabulary, a taxonomy, a thesaurus, an ontology, or some other abstract knowledge-organization system — is that we can make the activities of the enterprise significantly more effective or efficient by abstracting concepts from unstructured information, organizing that abstracted information systematically into a “schema,” and associating elements in that schema with documents.

While that may seem entirely laudable, it is not — at least not as currently modeled. We’ve been seduced by the convenient proximity of structural metadata (which is essential for retrieving documents) ... and by the omnipresence of the bibliographic cataloging model. Most pernicious of all, when we immerse ourselves into developing and applying knowledge-organization schemas, we become enchanted by the elegance of working in a different, higher plane of abstract knowledge, as if we were dealing with the heavenly instead of the profane. Unfortunately, the unstructured information itself still just lies there, like a bored prostitute, while ... Well, let’s not finish that metaphor.

The unspoken assumption that there can be no alternative to this bipolar relationship between the orderliness and refined economy of semantic metadata and the messiness of unstructured content is unwarranted. (In document-management and content-management technologies, addition of metadata is often an integral part of the document-import/management process.)

The assumption is, in fact, harmful because it is limiting. And some of our basic ideas about classification of content may be so flawed, so colored by unsupportable conventions and perceptions, that they threaten our openness to new, well-grounded models that help us integrate and leverage information in enterprises successfully.”

Jordon Frank, Traction Software

The answer is both. Free text does a good job but, here are a few reasons tagging is necessary:

  • Some content needs to be categorized intentionally. Easy example, I post something that is a piece of Collateral. I may not describe it this way in text, but may need it to show in a section on my website or blog (in case of Traction) as Collateral.
  • Tagging allows people to mark things as important, and therefore add to search relevance or improve navigation metaphors.
  • Tagging allows people to track things and support basic content workflows.

Bibliography and Links

Check out Bill Ives blog where he has interesting commentary on tagging at:

and social tagging: