Stepping into the kitchen now, Google introduced a recipe search engine Thursday that serves up dishes based not only on the ingredients you may have on hand, but the calories you want to consume and even how much time you have to cook.
About one percent of the queries on Google are for recipes, the company said, and the new vertical intends to make it easier for people to find recipes across the web. Keeping with its search roots, Google isn't showing the recipes in the results, and is linking to recipes on sites like Epicurious and the Food Network, which likely alleviates some worries about Google's newest effort.
The Recipe feature will show up in the left side bar, underneath other so-called vertical searches such as Shopping, News, Images and Videos, and gives searchers options to refine results: such as finding a red curry recipe, then specifying one that has less than 300 calories and includes red peppers.
But the real innovation is in the background: the entire search is built on structured data that webmasters have built into their webpages using markup code that's invisible to humans but is extremely useful to machines. The dream of the so-called semantic web is built upon the idea that web pages will be full of such underlying tags so that search engines can parse a webpage to learn someone's e-mail address or know exactly what a restaurant's operating hours are by scanning underlying code invisible in the browser.
That's a tough problem with the current web, according to Google's Jack Menzel, the company's product management director for search, despite the apparent ease that Watson had besting human Jeopardy opponents.
"We are still grasping for the Holy Grail of natural language search," Menzel said. "We take the approach that the internet exists, and it is so big and Wild West-like that you have to take it for what it is. It is this giant immutable thing that will do its own thing, despite what you want it to do."
The dream of a structured web has proven nearly impossible to create in practice as it requires coordination on building specs and then that web page builders take the time to mark their pages up in complicated XML. A more grassroots effort, known as Microformats, has had more success by focusing on just a few kinds of data and making innovative use of HTML, the lingua franca of the web, to simplify publishing meta-data. Google introduced its own suggestions of how websites could start publishing Google-friendly meta-data in 2009 (such as how many stars a rating is), with its so-called Rich Snippets.
And now for the first time, a mainstream search engine is built entirely around webpages that use microformats and other structured data.
So for instance, Google is able to show a searcher only Pho recipes that use tofu that take less than a half an hour to make, not by searching for pages that include the word "Pho" and "Tofu" and "Recipe", but by actually knowing that a recipe for something called "Pho" has an ingredient "Tofu" and a listed cooking time of 1 hour (for example, the is done after publisher's wrapping the word "1 Hour" in a defined HTML tag ()and then interpreting that in the search results ).
Much of what's held up publishers from including this kind of data is a chicken and egg problem. Why go to the trouble to add this underlying structured metadata to a webpage if it's not of use to anyone? And if there's no structured metadata, why would any company spend time and money trying to find it on the web?
Menzel estimates that only hundreds of recipe sites are currently wrapping their published recipes in meta-data, but they include biggies like the Food Network and Epicurious (which, like Wired.com, is owned by Condé Nast). And he hopes the recipe search engine will lead developers to create easy ways for every day bloggers to do the same.
"Our intent is to make better user expereince to see if we can jumpstart this ecosystem," Menzel said. "That way when someone thinks 'Hey, I just invented a great recipe, let me put it on my blog,' and that person's recipe should be a candidate."
But Menzel insists it's got to be easy and that Google doesn't want to push busy webmasters to do any work that won't result in more traffic.
"This is really a pragmatic response to the dream of the semantic web," Menzel said. "We would love if the XML world existed -- it would make search awesome, but no one is going to to do it. But we need to start somewhere, and a lot of the internet is built manually by people and their time is valuable."
See Also:- Microformats Go Large
