Is this a reasonable statement to make?
- Tagging is the process of adding descriptive terms to an item, without the constraint of a controlled vocabulary
- Folksonomy is the aggregation of tags from one or more users
Yes? No? Discuss.
(Full disclosure: You're helping me prepare for a tutorial on folksonomies that I'm presenting at the CSCW conference in Banff this weekend.)
Is a folksonomy an aggregation of tags from one or more users around a specific item/thing/entity? i.e. a given link on del.icio.us has a folksonomy?
The way it's phrased now sounds more like a tag cloud than a folksonomy, imo.
Looking at it again, I'm sure that's what you meant, but I didn't explicitly get the carry-over of the second one referring to the first in terms of a single item.
Actually, I've always considered a folksonomy to be an aggregation of tags for multiple items, whether from one or more multiple users.
Really? Maybe I need to do more research on folksonomies. It just seems like, without something to describe, what good is an aggregation of tags from a bunch of users?
What do you call the set of tags from a group of users on a specific thing? just its tags?
I want to go to your talk ;)
I'm claiming that a folksonomy can be:
a) multiple users tagging a single item (e.g. the del.icio.us bookmark example
OR
b) multiple users tagging multiple items (e.g. del.icio.us or flickr popular tags)
OR
c) one user tagging multiple items (e.g. your own tag cloud on flickr or del.icio.us [this one is debatable, based on whether you think a folksonomy requires a social component]
BUT NOT
d) one user tagging one item
Sounds like an interesting conference. Here are some thoughts, don't know if they'll help...
Tags are just keywords applied by users that are aggregated and exposed for navigation. To me, the social component is the distinguishing feature of a folksonomy. So, I would argue that folksonomy is not the aggregation of tags from one user, they must be aggregated across multiple users. Additionally, users must be able to see and be influenced by others' tags in the creation and use of their own set of tags.
The question of who is tagging what content is a separate issue. This is where Thomas Vander Wal's broad and narrow folksonomies comes in [1].
I've been thinking about tagging systems on a grid comparing private vs. social (tags vs. folksonomy) and own vs. others' content (narrow vs. broad).
private/own = google docs?
private/others' = gmail
social/own = flickr
social/others' = del.icio.us
So while gmail uses tags to organize messages, there is no folksonomy.
[1] http://www.personalinfocloud.com/2005/02/explaining_and_.html
I agree...I think the social aspect builds a folksonomy - letting others contribute to a shared classification of something. I think that's why Google terms them labels, because it's not a social, shared categorization influenced and consumed by others, just for your own use.
I'm a comment fiend today - makes me miss the LSC. I love discussions like these.
David - I often put the caveat around social when talking about folksonomies ("they are often social") as most social tagging services these days have social components. Many people also are tying their tagging in a del.icio.us type tool to their tagging in Gmail and in their OS (desktop or laptop).
Additionally, you may want to look at my recent presentation on folksonomy, Folksonomy for IA (http://vanderwal.net/random/entrysel.php?blog=1877). The Folksonomy Definition and Wikipedia (http://vanderwal.net/random/entrysel.php?blog=1750) is often linked to and quoted. While the Explaining and Showing post of mine at Personal InfoCloud is widely quoted, the narrow folksonomy is really not that valuable (it is better than nothing, but no where near as valuable as the Broad - or Folksonomy Triad). The past 18 months or so of experience since I wrote the Broad and Narrow post has proven this out. We do not get trending of usage of terms on one object (this is really valuable) and we do not get the ability to verify a group of people's usage of terms on objects (even more valuable). The collaborative filtering is also not available in a Narrow approach.
Liz - I fully agree that folksonomy is not one person tagging one item. The folksonomy is an expression of a person's vocabulary as defined by their tagging objects. We can watch a vocabulary emerge as they refine, expand, or modify their usage as well as the terms they use.
A folksonomy can be used to validate terms in a taxonomy, identify gaps in a taxonomy, and offer the terms that can fill the gaps.
Regarding one or more items, I often find looking at just one item is a valuable starting point when looking at the various terms that have been applied. I can identify possible trends, but these trends can only be validated by tracking a person's used of the term across multiple items. A collection of people using the term rather consistently across many items is a really good validation of a trend. Looking at tag term combinations is also interesting across objects being tagged. It is from this pairing of person, tag, and object that we can verify consistency and start using these for recommendation systems as well as identifying gaps in our taxonomy.
To be sure, the folksonomies I think of most often are built from tags, but is tagging a requirement? I think diggs, citations, URLs, and links can be the basis for folksonomies as well.
So what definition did you finally settle on to use at the conference?