Our work on automatic metadata generation is gaining recognition:
- Wilbert Kraan wrote about it for CETIS, under the nice title of “Look ma, no metadata forms“
As a reminder: you can play with the software on our site, build applications that rely on the web services that run on our web server or download the source code from sourceforge. Enjoy, reuse, and share!
This Wired News story is right on target: as search facilities become ever more powerful and usable, folders are bound to disappear!
It is more than a year now that I no longer maintain folders for my email: I only have an Inbox and a “read mail” folder. (In fact, that is not true, I start a new “read mail” folder every three months, in order to avoid more than a gigabyte folders. Then again, I also mirror all my emails to my gmail account and there I only have one folder with ALL my email.)
I started doing this when I realized that the overhead that went into maintaining the folder hierarchy didn’t weigh up to the way it helped me find things… Often, messages ended up misfiled, or they could belong to several folders, or the folder hierarchy no longer seemed appropriate, etc. etc. Basically, folders don’t scale to the number of email messages (or documents in general) that most of us cope with on a daily basis.
Early on, I relied on simple tools like Bloomba (since acquired by Yahoo) or NEO to search rather than organize and quickly realized this was a BIG timesaver. Nowadays, I rely on Apple Mail and spotlight or gmail.
And now this trend is accelerating and it begins to be more and more possible to follow a similar approach for all your files. I have long advocated doing a similar trick to learning object repositories. And then add some visualization, and realize that files are so 1960s and we can finally leave the desktop metaphor behind and start making real progress!
Anyway: if you’re still organizing folders, then why don’t you take a deep breath and try search instead? I’d be interested to hear about your experiences!
The Semantic Desktop is an initiative I hadn’t heard of before. It’s not very clear whether it is only a workshop at the Semantic Web Conference, or a pooling of efforts of initiatives like Haystack, Gnowsis and Chandler.
In any case, this is very much going in the right direction. Of course, metadata are essential to make this work and much of the metadata are collected from or generated off already available sources. I’d love to see more of this work and I’d love to see it well integrated with external information sources…
Friendfeed
Recent Comments