i202 Fall 2010 School of Information, UC Berkeley

Haven’t you run across images online where you’ve wondered what a particular item in the image was? And then read the image caption to discover that it told you nothing about what you were wondering about. Wouldn’t it be nice if we could describe parts of an image that may or may not contribute to the overall message in the name of the file or the image caption? Well, now you can. Nick Bilton’s blog entry Stipple Seeks to Tag the Web's Images on the NYTimes BITS blog is about the introduction of a free service called  ‘Stippleit’ to online publishers, provided by a company called Stipple that allows publishers of images to ‘tag’ different parts of an uploaded image with pertinent information such as what the image part is, and where to buy the item tagged within the image.



I think this blog entry is a great “202 in the news” post because in essence, Stippleit is the manifestation of an ‘Organizing System’. While Stippleit provides the potential to tag all the images on the web and claims to solve the perceived problem of information vacuum around images posted on the web, I am ambivalent about whether it is a successful organizing system in general. It has value if leveraged by organizations who have an interest in selling certain consumer products or services and if used sparingly with some governing vision (for example, a flower site, that tags only flowers in each of their images), but who knows how it will end up being used over time. The user has a one time initial cost of adding some code provided by Stipple to their site before they can start tagging but once they have done this they can tag older images when they choose to, or tag new images as they are being posted. In this way Stippleit allows all kinds of web publishers (individual bloggers, reputable news organizations, flickr groups etc) to organize images. Besides organizing, the service contributes to other “202 qualities”, such as metadata, classifications and collections. For instance, tagging image parts will also influence metadata regarding images, because every bit of information added by tagging contributes to metadata, and can be mined, collected and searched for. 
 


This service will add another dimension to IO and IR capabilities for images, but it’s like anything else on the web - the information could be reliable (i.e: an implicitly trusted source like The New York Times) or not (i.e: Wikipedia) and it could easily end up annoying if used in excess. Imagine scrolling over an image and having a bunch of links popping on and off, while you try to figure out where your cursor is, so you can get away from all the link popping.