An short but clearly written introduction to faceted search.
It’s no secret that tagging/labeling is the new black in the online world. It’s not a criticism, I’m a fan, I just suck at it.
Hierarchical (Tree/Directory) structures are very much going out of fashion and I don’t expect that trend to change until the great tech wheel turns again. The advantage of the hierarchical structure though is that it is simple – If you keep your categories broad it is pretty simple to find a pigeon hole that is “good enough” and we have all been trained to some extent by the Windows defaults of My Documents etc. Particularly great if you just don’t know where to put something you just need to save because once you decide where to shove it at least your other directories are unpolluted. This tends to create a directory that becomes the dumping ground until you are forced to re-evaluate the directory structure again. The limitations, however are rather obvious – files can only be stored in a single directory and once you get to multiple media projects it becomes an mess (storing a .wav in a sound directory and .jpg in a photo directory and the presentation in a “My Documents” directory) and difficult to move the projects around.
Tagging on the other hand is networked (many-to-many) relational meaning that things can be stored with multiple labels. Tagging makes it important to describe the thing when you save it which means that you have to think about what it is that you are actually trying to store – what are the different meanings that I might want to associate with this chunk of data? So you have a system that should (in theory) make retrieving the data easier right? Well only if you have a reasonable tagging scheme.
I am really bad at tagging. I either over tag (and end up with tags that are so diluted that they don’t help me retrieve easily) or too specific that after a few weeks I’ve forgotten. I think that tagging is an skill that you develop over time… I hope.