Retrievr is an interface which allows you to find Flickr photos by drawing a rough sketch (see the screenshot). It may not be as accurate as, say, a search on Google with words, but it’s also not bad. At one point, as I was attempting to elicit a red apple, a pumpkin was included in the results. I then tried to get the pumpkin back by sketching an orange blob with a short stem: I was unable to get a pumpkin included in the results, but I did get a bowl of spaghetti. It’s also important to keep in mind that it does not recognize sketches as iconic representations. For example, in the screenshot, Retrievr does not recognize my sketch of a rough green triangle with a short brown horizontal line as a “tree.” It recognizes my sketch in its entirety as a rough green triangle with a short brown horizontal line and a blue/green/white background (it took me a bit of experimentation to figure out the background piece of the equation: the background does matter when performing a Retrievr search). In addition, Retrievr currently only draws upon Flickr’s most interesting photos; Retrievr’s developers do invite submissions and a submission interface is in the works.

But enough about Retrievr’s limitations. I think that this method of search holds some promise for learners, as I said above, who may have trouble expressing themselves or processing information in words. It may also be a boon for those who simply prefer images and visualizations to words, especially if you combine Retrievr, Flickr, and, without trying to toot my own horn too much, my Greasemonkey script Trope for Flickr. Learners can start off a search by drawing a sketch in Retrievr and find the most appropriate Flickr photo(s). Because Flickr provides the ability and encourages photographers to assign tags to photos, a (albeit unvetted) descriptive vocabulary is provided for each photograph. With Trope for Flickr installed, icons are placed next to each tag allowing the learner to branch off to find information in other forms of media with the same tag, such as web pages (via del.icio.us), blog entries (via Technorati), podcasts (via Odeo), and academic references (via CiteULike).

Imagine the potential for access and power afforded by Retrievr to those learners who have been kept back from interacting with technology because the primary interaction mode of the World Wide Web is the written word.

The idea behind this photo group is to encourage as many people as possible from around the world to post one photo from one day. A collective snapshot of the planet Earth. In addition to being posted to the photo pool in Flickr, Michel and others have set up a A Day On Earth blog to reach a broader audience.

Some of the discussions were quite interesting and excellent in terms of fostering a community of practice. As a producer and participant in this community, here are some of the questions — beyond the technical ones (light source, composition, focus, etc.) that went through my mind as I searched for a subject:

• What story, idea, or concept do I want to get across?
• How do I want to represent myself and how I am feeling today?
• What is special about where I live and what I interact with that I would want to share with others?

It was not an easy task to address these questions, but what a great way to start the day. It’s a model that we should think about in the world of learning — deeply connecting with and reflecting on our world and those around us with the promise of a new day.