web2.0 notes

Monday, May 02, 2005

'Tags' Ease Sifting of Digital Data

New York Times
May 2, 2005, Technology

This is mainly about 'Tagging' tech, starting with digital photos, but it's been picked up by blogs for a while:

"Organizers of a blogging conference in Paris last week encouraged participants to tag their entries ''lesblogs.'' Italian blogger Luca Lizzeri did just that and got hundreds of additional visitors.
Sites like Technorati not only let you search its own indexes, but also pull items from other sites. So a search for ''tsunami'' brings together Flickr photos and del.icio.us links besides blog entries -- creating a mini-magazine of sorts on the fly.
Unlike hierarchical classification systems, taggers create categories spontaneously. There are no rules to craft on what categories to include and what falls under each.
Hierarchies ''are more accurate, but they move less quickly,'' said David Galbraith, founder of a tag-based wish list called Wists. ''It takes a long time for people to sit down and agree on them.''
Matthew Haughey, founder of the community blog MetaFilter, considered a taxonomy to organize archival posts but ''it's hard to make perfect categories and sub-subcategories.'' If you wanted to paint a fence, should you look under ''home and garden'' or ''household''?
So he went for tagging.
The blogging site LiveJournal plans to introduce tags in the next few months as an alternative to categories, and Rojo Networks Inc. launched a service last month for tagging news stories, so no longer are you limited to sorting items by publisher.
Of course, tagging has its drawbacks, and some Webophiles aren't quite convinced it will evolve into the Next Big Thing.
Consider classifications for a common pet.
''If one group decides we're going to call them `canine,' another `dog,' another `puppies,' ... when someone goes to search for what they call the dog, they are not going to pick up everybody's tagged instances,'' said Geneva Henry, executive director of the Digital Library Initiative at Rice University.
Engineers recognize the shortcomings and are working on better tools.
Search for ''automobiles'' of Flickr, and you're given ''cars,'' ''car'' and ''porsche'' as related options. Enough people tag photos both ''automobiles'' and ''cars'' that clustering software can tell they are related.
Another drawback lacks an easy solution, though. Once tagging takes off, marketers are bound to add irrelevant tags to hijack you to the latest Viagra ad.
Warns Danny Sullivan, editor of the online newsletter Search Engine Watch: ''The noise and deliberate manipulation will probably just bring the system into a crashing halt.'' "

0 Comments:

Post a Comment

<< Home