Felix Jung, a clever creator at avovision, has developed this slick Flash-based interface that uses Flickr's API (Application Programming Interface) to "grab" images based on Flickr tags to coincide with a song called "Astronaut" by Dan Frick.
Here are a few things Jung had to say about his creation: "Listening to Dan's song, the speaker makes me think of someone fairly isolated - far away either due to distance or time or both. When thinking of Flickr, I think of a multitude of people, all of them interacting with one another - sharing photos, comments, and memories. I liked the juxtaposition of someone singing about solitude, but having his words represented by... well... by everyone else. By the rest of the world.
Each time the Flash file is loaded, new images are randomly pulled from Flickr. I've hard-coded 53 keywords at set points in the song, and when the page is first loaded... calls are made out to Flickr to retrieve these keywords. With each call, I vary the parameters a little bit.
Let's say I search for the word "astronaut." In searching, I randomly apply a sorting method (date posted ascending, date posted descending, date taken ascending, etc). This randomized sorting method allows for different photos to show up in my results (each search should result in about 50 matches). From there, I randomly select one image, add it to the queue, and move on to the next word.
In some cases, I've taken a few liberties with my searching. Instead of sticking exactly to the lyrics, I've substituted words in certain places, either to elicit an effect or due to the fact that the word itself wasn't returning enough results. I've tried searching all text, but found that searching specifically for matching tags proved the most accurate." (You can read Mr. Jung's entire blog posting about his project here.)
I think there are some amazing exhibit and exhibition possibilities for museums and multi-media artists to make use of this approach to tap into existing public Flickr images (including those currently posted by museums.)