If you liked Mad Men, you’ll love this. It’s another one of those crowdsourced, digitized ephemera metadata projects—this time for 1960s advertisements published in the New York Times.

The project, put together by The New York Times Research and Development Lab, is called Madison (of course), after the New York avenue where many ad agencies were located. It uses the more than 150-year-old collection of digitized New York Times issues called TimesMachine, which the paper launched in 2008. (Incidentally, even then, getting a chance to look at old ads was cited as a major benefit to the digitization project.)

“While news events and reporting give us a glimpse of one aspect of our past, the advertisements that ran alongside those news articles allow us a very different view,” the Times writes in its description of the project. “They act as commentary on technology, fashion, economics, gender relations and more, often in ways that are fascinating, funny or poignant.”

But as we’ve seen with pictures, maps and menus, and other imagery, it’s difficult to develop optical character recognition (OCR) software that accurately discerns their meaning. “Although the software is good at indexing news stories, it’s not as good at recognizing ads—which, in cultural terms, can be almost as interesting as the actual journalism,” notes Mathew Ingram at GigaOm. Moreover, ads often have complex layouts and elaborate typefaces, which makes them more difficult to scan for text, writes the Times’ creative technologist Jane Friedhoff.

That’s where the crowdsourcing comes in.

Much like the New York Public Library is doing with its maps and menus, the New York Times is asking people to look at a series of images that its software believes is an advertisement, and then asks them to perform various tasks on the imagery. A contributor could choose to simply “find” ads by identifying whether the image is an ad (or, perhaps, multiple ads, or part of an ad), “tag” advertisements with a category and company names, or do the extra work to actually “transcribe” all the text in the ad. The Times Research and Development Lab actually put quite a lot of thought into how to divide the work into tasks simple enough to be done easily and on a variety of devices.

The New York Times archives are full of advertisements that give glimpses into daily life and cultural history,” the project’s splash page explains. “Help us digitize our historic ads by answering simple questions. You’ll be creating a unique resource for historians, advertisers and the public—and leaving your mark on history.”

Even without this new Madison initiative, the TimesMachine project alone demonstrates that as far back as 2007, when it first started, the New York Times was pretty leading-edge in terms of digitization. The digitized paper from 1851 through 1980 consisted of 11 million articles, taking up 4 TB, the Times wrote when launching the project. The R&D Lab itself consists of eight people on the 28th floor of the Times building; these folks are charged with building things based on what the lab predicts for three to five years in the future, writes Advertising Age. Similar Lab projects, more text analysis-based, are Chronicle, which tracks the use of words over time, and Curriculum, which tracks topics in the Lab employees’ browsers.

The Madison project is also the first built on Hive, which the Times describes as “a modular, flexible, open-source platform we have developed in the R&D Lab on which any number of crowdsourced applications and tools can be built.” (It is not related to the Apache Foundation’s open-source software by the same name.)

The Times plans to send out one decade’s worth of ads at a time. Naturally, playing off the success of Mad Men, it picked the 1960s first. As we all know from watching the show, the 1960s were a time of great experimentation in advertising. So in the Times, people are discovering such gems as ads for the Volkswagen Beetle, Woodstock (including a coupon you could clip and mail in to get tickets—for $7), and a help wanted ad for the government agency to build what became the Internet.

What’s particularly interesting is the speculation around how project contributors might be rewarded for their participation. Organizations such as GigaOm and NPR have wondered whether such projects might presage a future where people do this work to receive some sort of membership-based project, such as free or discounted subscriptions.

But while the Madison project does include gamification elements—you get rated as a “reader,” a “finder,” and so on depending on how much work you do—the real reward is intended to be getting to see the ads themselves. And that’s deliberate, Friedhoff writes—building off work that the New York Public Library did with its menu project.

“We also purposefully chose an approach that downplayed gamification in order to place the fun part of a potentially dry process at the forefront: discovering and sharing interesting cultural items and artifacts with your friends,” she writes. “Rather than trying to tempt a user into participation with external, material rewards, we aimed to design a system whose biggest rewards came from engaging with it: namely, discovering and sharing a piece of culture that probably only a handful of people have seen since its original publication.”

The advantage of doing it that way, she continues, is that people are less likely to game the system to get the rewards.

And it appears to be working. “You never know what's next,” writes Amanda Kooser for CNET. “It might be a slew of classified ads offering up office space, large visual ads for department store sales or exclamation-point-packed ads for dune buggies, sand chariots and a weird all-terrain vehicle called the Terra Tiger.”

What will the rewards be once the Madison project is complete? Besides obviously being able to see a bunch of cool ads, it’s possible that the ads themselves could become the basis for future retro ads, Ad Age hints, noting that a similar retro effort for Newsweek—tied to the fifth season of Mad Men—was successful.

Related Posts