Attaching meta-data or not

Dave Winer points out in some comments here that what we was really trying to say was:

You guys want users to enter metadata, I’m looking for ways to get around tha because I have found that people don’t even spell things right, much less label things.

He’s right- of course.

Nobody wants to attach tags, categorize, define their posts as something specific or even THINK about what theyr’e doing - first.  It’s a “shoot from the hip” socieity.   Paul Kedrosky calls them lazy.

Both Paul and Dave are right.  So what do we do?

The ‘automatic’ school of thought believes that your content itself is enough to discern what it is you’re about.  Others think this knowledge can be discerned from who your Friends are or what Groups you’ve joined.  Or what music you listen to.

My favorite automatic technique is simply matching the interests, friends and groups you’re a member of - to each other - and viewing those similarities.  You can (at least) then become a FRIEND of someone who has similarities with you.  That’s what we did at 1UP.com.

So is the issue: “automatically figure who people are - so they don’t have to manually attach meta-data” or is it something else?  Is it really such a chore to attach tags?

Well as Dave points out - I don’t categorize or attach tags to my posts - so call me lazy.

Do we really think we don’t NEED tags, folksonomys or even meta-data at all?

HHHHHHHHHHmmmmmmmmmmmmmmmm - I’m not really sure.  I don’t there’s one answer. People are different.  Clearly Dave Sifry and Dave Weinberger like tagging.  Me - I kind of liked Matt Mower’s “Live Topics” and Paolo’s “k-collector”. And anything that Phil does.

BTW I did notice that Paul puts his favorite books on his blog. I guess he’s not too lazy for that.  And Dave Winer - well this is not a man anyone would call lazy. He categorizes, plots, designs and gets analk - just like the rest of us.  Only in his own way.

I doubt any algorithm could figure out what Dave is about.

 

3 Responses to “Attaching meta-data or not”

  1. Phillip Pearson Says:

    I think Dave is saying “forget about the metadata, just look at what you WRITE about”.

    Simply looking for keywords in blog posts would get you a long way towards finding what people are interested about. Links and tags would be next.

    Then, if you’re Danny, you dump all this out as RDF and feed it into a chain of tools written by RDF hackers… while Dave would push it out as OPML and evangelise the hell out of it until someone builds it :-)

  2. Danny Says:

    What Phil said.

    Remember it’s not either-or, you can use text processing tools to extract information alongside tagging etc. Explicit metadata (like the date of RSS items) can boost the collective value of all the metadata available, including the text-mined stuf. Although manually entered material is the most expensive for the user (e.g. the creator of an RSS feed) it’s potentially the most valuable, and if the systems are built well the return from it can be maximised.

    Once you have that information, as Phil hints there are plenty of choices on what you do with it. Yup, I’d use the RDF/OWL stack, it’s designed for processing resource descriptions on the web.

  3. card games online free holdem poker online Says:

    card games online free holdem poker online

    characteristically denotable collaborate sloped blurting duplication free holdem poker online tournament http://www.theholdempokeronline.com/free-holdem-poker-online.html