Josh Porter has asked on his blog what practical value people can get out of folksonomies, and Steve Rubel is trying to come to grips with potential marketing implications. A folksonomy is just the aggregate of tags (labels) people have applied to a resource in a shared archive. The best known example is del.icio.us, a system where people label and store their browser bookmarks on a shared web site. The tags for these bookmarks can be thought of as one-word descriptions of what the bookmark is about. The folksonomy for a particular bookmark is the collection of tags the various people who bookmarked it have applied to it. So people might tag the New York Times book review site as variously “books”, “literature”, “idiotReviews”, etc. That collection of tags is the folksonomy for that site. The word folksonomy can also be used for this same group classification process applied over many sites.
So, what's the practical value of all of this? My cut is that one of the main sources of value is in tracking the spread of memes (e.g., pivotal ideas, advertising slogans) that can be summarized in a word or short slogan. Are people picking up on a particular way of viewing the world, perhaps one that you espouse?
There are examples of businesses doing this already. For instance, just today, Richard MacManus of Weblog Solutions Ltd. (a New Zealand weblog consultancy) presents a study of how people have, in aggregate, classified two web articles. He considers these articles as pivotal examples of a new concept he is trying to promote, Web 2.0 (“the web as platform”, see his article). His method was to go on del.icio.us and do a query to obtain all the tags people were using when bookmarking the articles. The short of the analysis is that he discovers no one is using “web 2.0” (or at least very few in aggregate).
His response to this lack of uptake is to put out a better definition of his concept to try to get people to adopt the term for every day use. He also wonders if he should continue classifying his posts as “web 2.0” since the term seems so little used. His response is that since noted people are using the term, it should become more widespread, and he wants to ride the wave.
Jon Udell takes a slightly different perspective. He looks at an individual URL and sees how many times it is bookmarked by day over a given period to attempt to establish trends in the URL's uptake. The idea is that the content at the URL expresses a certain idea and that bookmarking it expresses some sort of cogent reaction to the content (most likely agreement, but not always).
In using this approach, Udell seems to be implicitly agreeing with a post I made a couple of days ago about how the main value of folksonomies might be in discovering other people who were looking at the same resources you were because then you knew you at least had something in common.
Thinking across Udell's and MacManus's approaches, it seems that Udell's has the advantage of being more general. He actually used two sources, del.icio.us which supports folksonomies and bloglines which does not. However, by not looking at the individual classifications made in the del.icio.us folksonomy, Udell is losing the ability to track the preponderance of the meme, that short one word description that indicates that a worldview has really caught on. A weakness of both approaches is that they limit themselves to the digital cognoscenti. Del.icio.us has thousands of users, and only a very small proportion of web surfers use a web site like bloglines that allows you to essentially do all of your blog and other syndicated content reading in one place.
One might wonder why you would not just use Google pagerank (a number ranging from 0 to 10 that indicates the extent to which the search engine has determined the page you are viewing is highly regarded by other web surfers) or some other metric as has become so commonplace in search engine marketing. For one, Google page rank is a blunt instrument. It does not provide the level of detail needed to do either one of these two analyses. Two, Google page rank is constantly being gamed, making it a less reliable indicator of relevance. However, the relative cleanness of folksonomic systems may not last long. Clay Shirky is already reporting folksonomy SPAM on del.icio.us
Leave a comment