World’s Most Abstract Visual Dictionary Defines Words By Their Top Google Image ResultBy: Josh Wolford - July 5, 2012
The idea of a visual dictionary isn’t new. Hell, I can still remember exactly how my first dictionary looked, felt, and even smelled. It was a kid’s dictionary and contained a lot of pictures – enough to help a young, aspiring word nerd.
But this visual dictionary may not be suitable for kids – that’s because the images that comprise it come from the interwebs.
Designers Ben West and Felix Hayes have created “Google,” which is the world’s first dictionary that uses imaged pulled directly from Google image search.
“If the internet goes off, you may need this reference book Felix and I made,” says West on his site. “It contains the first Google Image for every word in the dictionary.”
Yep, all 21,000+ of them. The dictionary comes in at a staggering 1240 pages of internet images, all selected because they were lucky enough to be result numero uno in a simply Google search.
“We used two PHP scripts my brother Sam wrote for us,” said West in a statement to Creative Applications. “The first one takes a text list of dictionary words and downloads each image in sequence, and the second lays them out into columns and outputs a PDF.”
Apart from any sort of functional uses to the Google Images-baed dictionary, is it an artistic statement? Is it just a curiosity? One thing it is for sure is an incredibly interesting look into Google Image search and just how accurate that first picture really is.
For instance, a quick search of about ten tangible items like “fork” and “mint” produced great results. The first image that popped up was about the best physical descriptor of the word in question. Score one for functionality.
But searches for less tangible, more abstract words produce a mixed bag of images. Start searching emotions and heady verbs and some fo the first images can be a tad misleading. Or maybe just non-specific. The point is, the “Google” book will be pretty abstract at times, to say the least.[via TechCrunch]