Digital Reading: Coming to a Cookbook Near You—The Semantic Web and Open Data

Anne KostickBy Anne Kostick, Partner, Foxpath IND

Last month, The New York Timess Julia Moskin wrote about the difficulty of finding good cooking recipes on the Internet. The article had the well-optimized title, “Can Recipe Search Engines Make You a Better Cook?

A follow-up question all cookbook publishers should be asking is, “Can recipe-search technology help create a better cookbook?” They should ask quickly, because the print cookbook is close to being obsolete.

Cookbooks are already a fairly modern aggregation of curated information. A cookbook reader is likely to find images, instructions, cross-references, basic and technical information, elegant narrative, and history—oh, and recipes—all in one place.

But what’s happening with recipes and food information on the Web is changing cookbook customers. Cooks, readers, and food enthusiasts are getting used to a higher level of information service, not to mention timeliness, in their cooking lives.

And now we begin to see the fruits of the intersection of technology and food (a busy intersection, by the way: techies are often foodies, and vice versa) with everything from (to name just a few) the overachieving, niche-y Modernist Cuisine by Microsoft pioneer Nathan Myhrvold, to the powerful iOS app, How To Cook Everything On the Go from Mark Bittman and apps from the Food Network and Martha Stewart Living.

Here’s a category where the question “Is it a book or an app?” is irrelevant. These cookbook versions can have the ability to search and filter recipes to a very high degree; cross-reference recipe ingredients to basic information about ingredients and illustrated technique tutorials; and draw from a long list of useful and mouthwatering features.

Underlying these abilities and enhancements is the Semantic Web —technology that enables computers to read and understand the relevance of data in order to search, filter and recommend the best results for any query. These methods power Google’s Recipe View (the journalistic interest in Internet recipe search was sparked by Google’s recent release of this new feature that allows users to search only for pages with recipes on them) and many other initiatives, such as the searchable recipe aggregator, and the recipe-formatting and databasing platform from Epik.

Parsing a recipe, from

Understandably, developers in the Semantic Web have a fondness for open data—huge accumulations of information on every possible subject that is available, free or for a low fee, to anyone who can use it in a website or app. Food data is particularly rich for Semantic Web development, and is particularly connectable to open data sources.

If your cookbook project could benefit from a high-quality database of ingredients with all their properties and uses, you can get one. If you think the book’s recipes should convert automatically to metric measurements upon request, it can happen. If you want to incorporate a shopping guide that connects you to local produce in season and prompts you to cook recipes that use those items, it’s as good as done.

The next generation of cookbooks will be—must be—more than print products followed by text-and-picture digital conversions. Digital-born cookbooks will span the gap between reading and doing while still presenting beautiful photography and illustration; lively narrative and commentary; and the well-crafted (and now highly searchable) recipes of an expert, a TV star, or a talented teacher. They’ll offer a real-time, filtered connection to the best information the Web can supply, in the container of a curated digital book. Who wouldn’t want that?

Moskin’s article followed up on an earlier Diner’s Journal entry, “Is Google Making You a Bad Cook?” and accompanied another Diner’s Journal entry, “Google’s Most Wanted Recipes.” Clearly, the Times is concerned about its readers’ level of culinary accomplishment and is worried that cooks are accepting a lower quality level in their crowd-rated and keyword-searched recipes. Happily, many online commenters replied that they were immune to the flood of crowdsourced recipes thrown at them by Google because they relied on cookbooks.

But consider this: in the iTunes store the How to Cook Everything app is $4.99 for the iPhone and $9.99 for the iPad; the HTCE iBook is, necessarily, $22.99.

Reader, I bought the app.

NOTE: DBW has an Editorial Forum on LinkedIn, a sub-group for editors and others working in trade publishing to discuss standards, workflow, best practices, and the general Qs that most print people feel when confronted with terms like “workflow.” The Forum is moderated by Anne Kostick and David B. Schlosser. Anne’s weekly column, Digital Reading, discusses the field of User Experience and explores what it offers to trade publishers.

Anne Kostick is a partner in Foxpath IND, a digital-print-web consulting and services company specializing in the transition to and from traditional content development, management and publishing. She is also the editor in chief of Dulcinea Media, an online publisher in the educational market, and is the current president of Women’s Media Group.

3 thoughts on “Digital Reading: Coming to a Cookbook Near You—The Semantic Web and Open Data

  1. Yvette Chin

    It’s funny: digital cookbooks have been on my mind a lot lately because I recently picked up the America’s Test Kitchen / Cook’s Illustrated iPhone app (more on that in a minute).

    Having played w/’s amusing “recipe spinner,” I never really gave much thought about how it works until you mentioned semantic web:

    I’ll admit: the spinner makes it so that you can choose pretty outrageous things “Lamb-based beverage made in a slow cooker, anyone?”, but that’s part of the draw. Of course, the difference is that allrecipes draws entirely on its website: crowd-sourced, natively online, already marked up, dynamic — all things that traditional cookbooks are not. How do we get there? Do we want to? No easy answers, of course.

    But really, I wanted to talk about America’s Test Kitchen: here’s a PBS show that has both a print mag (Cook’s Illustrated), print and e-cookbooks, and an app. While I’ve watched the show off and on for years, I only subscribed to the print mag a few years ago (my main method of consuming this brand), and I only discovered the app a few weeks ago.

    A few weeks ago!? That really says something about problems of discoverability in the face of years of brand loyalty. Yet, I imagine that they are doing well, extending the brand into so many platforms beyond the original PBS show. As for the app itself, it’s working on the same freemium model that ATK uses with its website. Print subscribers get VIP access to premium content on the site and on the app. That said, there’s a lot more available on the website, than on the app.

    To their credit, I think ATK is really extending its brand into its app the right way. This app isn’t really about bells and whistles (although the in-line timers are pretty interesting/useful), it’s about the ATK method: describing recipes or taste tests in a way that shows the completely scientific method that goes into testing each dish. It’s the same kind of prose and method; they seem to be staying true to their mission statement.

    But this is where the story goes south: even now, looking at this app, I can sometimes see the HTML tags… a crack in the armor for a brand that I really found no fault in… until I looked at the app….

    1. Anne Kostick

      I would ask if ATK is incorporating semantic web search and using open data to expand the ultimate usefulness of their digital products. That’s the exciting direction. Without it, I think individual apps or digital will be seen as just like another shelf-load of cookbooks with no meta index or discover-retrieve capability. Cooks will soon come to expect the cookbook to understand the difference between all-meat chili and bean-chili (to use Julia Moskin’s example) and return only the relevant results…and that’s what can happen.



Your email address will not be published. Required fields are marked *