While the term mash-up has its roots in hip hop culture, the web mash-up seems a natural evolution of our need to customize and our love of hacks. (See an earlier post on creating “clever solutions to an interesting problem.”) In Sampling the Web’s Best Mash-Ups, Business Week provides “a guided tour of the some of the most innovative sites that combine data and features of other sites to make something new.”
Here are most of the highlights:
One of the most useful mash-ups, 3D software engineer Paul Rademacher’s creation was born of frustration. He needed to find a house in Silicon Valley for his job at DreamWorks Animation and grew weary of the piles of Google maps for each and every house he wanted to see. So he created a Web site that mashes up listings from the online classified-ad service Craigslist with Google’s mapping service. Choose a city and a price range, and up pops a map with pushpins showing the location of each rental.
“I wanted this to be a real application, not just a Web page,” says Rademacher. “The key will be what’s useful, not just what’s cool.” His site has spurred a rash of Google Maps mash-ups, including some of the following slides.
An experiment for “exploring a place, based on the photos people take,” this site places photos on the Flickr photo-sharing site onto a map of the U.S. and Canada. Created by Eric Rodenbeck and Michal Migurski of the San Francisco Web site and applications design firm Stamen Designs, Mappr! uses the tags, or phrases, that are customarily appended to Flickr photos and analyzes them to infer where they were taken. Type in, say, “Route 66” and see the photos strung out along the famous highway.
“Mappr! doesn’t know anything about Route 66, and Flickr doesn’t know anything about Route 66, but together they found patterns that didn’t exist in either one,” says Rodenbeck.
If you ever wonder whether to search on Google or try Yahoo!, you no longer need decide. Computer-science grad student Ashish Gupta mixes search results from both sites into one list at DoubleTrust.net. Not only does the site return page results that are ranked high by both search leaders but it offers lists of “orphans”—Google results specifically not shown on Yahoo, and vice versa. And it shows 80 results per page in compact form, far more than either Google or Yahoo.
Strictly for fun, Francis Shanahan’s virtual photo-collage site isn’t exactly a mash-up. It’s what’s known as a remix, a site that creates something entirely new out of another site’s data. In this case, collages of famous figures, such as Amazon Chief Executive Jeff Bezos and actor Julia Roberts, are drawn from thousands of apropos pictures of Amazon product pages. A collage of author Stephen King includes product pages for Halloween costumes and a Hilary Duff video (which may be scary to some people).
“In the next 12 months, people will be thinking about the Web in a new way,” says Shanahan, a senior software architect at Sapient. “The folks who can combine these disparate technologies and services will do well.”
Getting into the mash-up game with its own one-year-old search site, A9.com, Amazon has created a site that essentially allows people to create their own customized mash-ups. Its new Open Search initiative lets hundreds of other Web sites submit data feeds that can be searched by A9 users. So searchers can make queries not just of the whole Web, which can return lots of useless links, but to specific sites, such as that of The New York Times or NASA, all without leaving A9.
Amazon hopes to spur more such mash-ups from outsiders. Says Amazon Web services evangelist Jeff Barr: “The more you’re able to unlock data and give it out, the more people can create these new services.”