Archive for November, 2007

Flickr and user-generated metadata: what is all this participation leading to?

November 4, 2007

Flickr has been making waves both online and among information professionals since its release in 2002.  (http://www.flickr.com/ )   I am sure many readers will have browsed Flickr’s content in recent years, or perhaps even have accounts with this popular photo sharing utility.  This review will focus on Flickr’s notable Web 2.0 innovations and some of the challenges and opportunities that result from large-scale user generated content rather than how to use the site specifically. 

Flickr has two goals at its center.  The first goal is to allow people to make their photographs available to “the people who matter to them”.  Flickr therefore allows people to choose the level of exposure they wish for their photographs to have.  While an aspiring photographer may make his photographs available to the entire web, another user may be more comfortable restricting her viewers to a carefully selected group of family and friends.  This kind of widespread sharing of personal information is clearly a large part of the Web 2.0 landscape, as we have seen in our discussion of the social networking sphere.

Flickr’s second goal is to provide new ways of organizing pictorial content.  Due to the exploding number of photographs in personal collections (thanks to widespread use of camera phones and digital cameras) the physical album is often no longer adequate for organizing such a large number of photos.  Flickr acknowledges that few people have the time to organize their vast collections of pictures, and so it proposes that an answer to this problem is collaborative organization.   Metadata tags, comments, and notes can all be ascribed to photographs posted on Flickr, and each of these is full-text searchable.

User-generated metadata is a very hot topic in library school right now, with opinions varying wildly.  Regardless of whether you believe that tagging serves order (by providing a scalable, flexible solution to the description of online content,) or chaos (as it uses no controlled vocabularies, synonym control, or standards,) tagging is a very popular function that Flickr users have widely embraced, and it is worth exploring.  The idea is that with enough users contributing descriptive terms to an image, over time a cluster of the most relevant words will emerge as many users independently ascribe the same terms to an item.  Tag “clouds” like the one displayed here, visually display the most commonly occurring tags within a system. 

 If you are unfamiliar with social tagging and folksonomies, I highly suggest reading this article by Clay Shirky.  Though biased, Shirky does provide a clear description of how tagging works, where it is being used, and some of its broader implications.  While I do not agree that folksonomies offer a complete solution to the problem of organizing online content, they do provide a fascinating glimpse into the descriptive terms that are most relevant to specific user groups at a specific moment in time.  Similarly, tags are continually updated as new terminology is invented and the connotations of various words change among certain audiences.

So, what happens when the popularity of programs such as Flickr generate hundreds of millions of freely searchable photographs online, many of which are coupled with descriptive metadata?  In essence we are beginning to see opportunities for the development of new applications to facilitate the discovery and use of visual information.   Recently, for example, Flickr launched its map function, which allows users to drag and drop photographs onto a map, creating a diverse landscape of photographs.  A quick search for “Flickr map” on Flickr’s home page will allow you to view users’ map applications to get a better view of what is possible with this feature, but I couldnt’ get permission to feature a specific user’s works in time for this post.  I highly suggest checking this function out in detail!

A much more powerful example of how vast content and user-generated metadata can be used in cutting edge visualization technologies is better illustrated in this video.  It is about five minutes long and describes how models of the earth can be created through aggregates of user-contributed photos to help people search and discover data on a grand scale as well as a very granular level.  This project all photos tagged “Notre Dame on Flickr were aggregated and placed within an interactive, 3-D model of the cathedral.  Please do have a look at the video.  This research hints, I think, at where we are headed thanks to Web 2.0 culture of sharing and participation.

 I would love to hear some of your comments on the video, Flickr, and especially user generated metadata.  I hope to hear from you!

Flickr cool rating:  5- easy to use, with a highly developed community and privacy options make this site tremendously useful and innovative.  Very cool!

 Additional Reading

Here is an article detailing the Brooklyn Museum’s use of Flickr to create interactive exhibits:  http://www.archimuse.com/mw2007/papers/caruth/caruth.html

If you enjoyed the video and would like to explore this research in more detail, please check out Microsoft Live Labs’ preview of Photosynth: http://labs.live.com/photosynth/default.html