ITags – something she and her colleagues are proposing to deal with some problems with tagging. She’s started a company called Dabble, which is a video remix community.
There’s a problem with technorati tags – people want to be able to publish their own stuff on their own sites and have people know that it came from their own sites. There are lots of different uses for tags and people want to participate in communities, but still publish on their own sites.
She did a usability study last spring, centered around text bloggers – what they asked for was the ability to create tags that didn’t require a link. At same time they realized their categories were being treated as tags in aggregation systems. Tags needed to be trusted – wanted to know that tags weren’t spam but didn’t need to know the actualy maker. Visibility vs. invisibility. Want to create their own tag clouds, and to tag objects separately from blog posts.
35% of technorati tags are actually tags, but 65% are blog categories. In creating Dabble they’re collecting media not just tagging it. They’ve collected 68,000 items so far – about half have tags. The richer the media, the more tagging you get. 10,000 of the videos come from independent sites with video on it.
What happens when people check out video is they look at the duration and the tags, much more than the titles.
What are itags – identity, creative commons license and an object. All of this can be bound to an object, which can move. The degradation of URLs is very fast, but the media lives on somewhere.
The identity piece – can be well-known, psuedonymous, or a blog url. You can protect the suer by obscuring who they are. When you trust tags (if you’re an aggregator or host) you have to trust the maker of the material – so you need some sort of identity. This allows freeing of media from file location to be site-independent – could use XRI as a structured identifier.