Folksonomies
I just stumbled upon a discussion about 'folksonomies', which is a concept that describes emergent phenomena like Flickr, Orkut etc. In these web services the users tag the data they enter, for instance by naming a picture in Flickr. Askpang refers to Thomas Valder Wal who describe this activity as a way of making bottom-up taxonomies.
Louis Rosenfeld explains that one of the downsides with folksonomies is that the informal tagging or categorization they are based upon, isn't very useful if the user wants to search for specific material, or need to navigate through it in more systematic ways. Controlled vocabulary (or strictly developed categories) that are made by professionals and applied by professionals, will of course be more coherent and much easier to manipulate and to perform searches on.
I think maybe that this 'ordinary people versus professionals' discussion is confusing two different cultural phenomena. In an emergent system like flickr, some of the joy is that the users can toy and improvise with categories, with how the information is structured. The result is that the personal artifacts or data from the users is weaved together with other users data in new and unpredictable ways. This largely describes how cultural traditions evolves; by being used, combined and changed over time (also much like memes are spreading). This type of emergence is distributed and bottom-up, but I am not sure if I would call it a taxonomy, since the tagging or entering of metadata isn't made for categorizing as such, but for making a specific set of data communicate with others of the same kind. Without any formal agreement of how the data should be classified, it becomes a matter of sheer luck if it is labeled or tagged in a non-conflicting and coherent way.
The other phenomena (or the professional indexing activity), is how information systems is being developed. To create a system like a library index or a newspaper archive, you have to tag all data in a uniform way. These systems are not made for improvising, and having the same rigid universal system is what make information retrieval efficient. These kinds of taxonomies are by nature top-down driven, given that the makers must have the wholeness of the data in mind, when developing the taxonomic system.
The reason why these systems work differently is because they are made for quite different purposes and are used in different ways. I therefore find it strange that people seem to regard them as competing, and discuss which one of them will "survive", given that folksonomies are much cheaper than professional indexing. I honestly don't think economic reasoning is very useful on this subject.
Louis Rosenfeld explains that one of the downsides with folksonomies is that the informal tagging or categorization they are based upon, isn't very useful if the user wants to search for specific material, or need to navigate through it in more systematic ways. Controlled vocabulary (or strictly developed categories) that are made by professionals and applied by professionals, will of course be more coherent and much easier to manipulate and to perform searches on.
I think maybe that this 'ordinary people versus professionals' discussion is confusing two different cultural phenomena. In an emergent system like flickr, some of the joy is that the users can toy and improvise with categories, with how the information is structured. The result is that the personal artifacts or data from the users is weaved together with other users data in new and unpredictable ways. This largely describes how cultural traditions evolves; by being used, combined and changed over time (also much like memes are spreading). This type of emergence is distributed and bottom-up, but I am not sure if I would call it a taxonomy, since the tagging or entering of metadata isn't made for categorizing as such, but for making a specific set of data communicate with others of the same kind. Without any formal agreement of how the data should be classified, it becomes a matter of sheer luck if it is labeled or tagged in a non-conflicting and coherent way.
The other phenomena (or the professional indexing activity), is how information systems is being developed. To create a system like a library index or a newspaper archive, you have to tag all data in a uniform way. These systems are not made for improvising, and having the same rigid universal system is what make information retrieval efficient. These kinds of taxonomies are by nature top-down driven, given that the makers must have the wholeness of the data in mind, when developing the taxonomic system.
The reason why these systems work differently is because they are made for quite different purposes and are used in different ways. I therefore find it strange that people seem to regard them as competing, and discuss which one of them will "survive", given that folksonomies are much cheaper than professional indexing. I honestly don't think economic reasoning is very useful on this subject.
posted by Faltin at 5:52 PM
0 Comments:
Post a Comment
<< Home