Four pillars: Musings on tagging

[Before I begin, I have to declare something, almost as a conflict-of-interest. Some years ago, I was talking to the guys (at work) who were really committed to semantic web concepts, and promised them I would never set up a project called semantic web. There’s something about the interplay between emergence approaches and detailed structure I find truly fascinating]

Niall Cook followed up on my last post re tagging and Cogenz, and pointed me towards the Lucent and IBM studies and achievements.

Very interesting. You can read Niall’s post here.

Somewhere in my head, I’m still not sure. You see, I view Flickr and last.fm and oodle as variants of the same theme, the disaggregation and reaggregation of search and publishing.

Here’s what I think:

  • 1. We provide people with simple tools to tag and post simple details about a “something”. It can be photos, books, music, cars, whatever. But it is ONE “asset class”.
  • 2. The “we” that provides the simplest and most convenient and easiest-to-use tool starts getting “liquidity” for that asset class.
  • 3. For a while it’s game over, because network effects rule and first-mover-advantage means something in this sort of digital scaling. Collaborative filtering makes it even stickier.
  • 4. Then we learn more things about said “asset class” and find ways to differentiate yet again. And new tools emerge and new liquidity points emerge.
  • 5. And then the cycle continues and we all enjoy ourselves.

All this works because (a) people know where to go (b) it is easy to add or remove things and (c) there are few, if any, format issues.

This format issue is no different from DRM in some respects. From my viewpoint they are one and the same thing. Where am I going with this?

Simple. When the asset class we speak of is “information” we have some real problems to contend with. I can’t just bookmark the things I want to bookmark and share with the rest of my community, even within the firm, because of variants of the format/DRM problem. Sometimes it’s called format. Sometimes it’s called authentication and permissioning. Sometimes it’s called image rights or intellectual property rights or even plain old copyright. And sometimes it’s called confidentiality or data protection or secrecy.

So while I like what I can see of the IBM and Lucent work, and I understand what is intended by Cogenz, my jury is still out. The market for socialising information as an asset class is still unformed, nascent, with some real problems to overcome. I can’t even provide sensible library/information services within the firm as yet. Tagging within the enterprise will help me solve this, for sure, but the impact will be marginal until and unless I can allow people to access more of the external sources of information sensibly.

But there’s enough in the premise to make me think. Can I create a skills matrix for a firm “at source” by asking people to post their CVs into a pool? And make it a simple profile, a microformat, with tags that enrich the meaning. And get over the confidentiality thing by asking the owner of the info to post voluntarily and without force. Where else can I do this sort of thing, avoid a format/DRM problem by transferring responsibility to the “beneficial owner”?

This CV thing, by the way, is another fossilfools thing I can’t get over. All firms hire people with specific skills and competences and experience and then make it quite difficult to find out about those skills or competences. Even the ones who believe in knowledge management and invest heavily in related systems. Something broken here.

Let me know what you think

This site uses Akismet to reduce spam. Learn how your comment data is processed.