A Blog by Jonathan Low

 

Aug 8, 2014

Data Dawg: Recipe Research and the Demographics of Detrimental Dining

"It aint the meat, it's the motion," sang Maria Muldaur, but these days, the big data corollary might be, it aint the info, it's the inference.

The more data we accumulate, the more opportunity it provides to extrapolate and draw conclusions about related behaviors. This can cut two ways, as those primarily concerned about health, for instance, want to know what the aggregated knowledge implies about customer preferences and the impact that may have on morbidity, mortality and the cost to society of indulging or circumscribing the associated inputs.

Others, more concerned about individual freedoms, are raising questions about how this information is used, especially if it is not voluntarily provided. Some maintain that civilization benefits from change imposed by superior powers. Others want to be able to buy bacon or beer or contribute to a political candidate and not have to worry about whether it will affect their insurance premiums, credit scores or overdraft privileges.

Seemingly innocent information, like the types of recipes one researches online can, as the following article explains, be used to make other assessments about health, well-being and actuarial expectations. While there is nothing wrong with the gathering and analysis of this data, perceptions about its potential uses are fueling anxieties that, if left unaddressed, could challenge the broader use of beneficial information. JL

Kevin Fitchard reports in GigaOm:

Yummly analyzed over a billion individual data points from 100 million user searches over the last year. Then it used its own recipe analytics to break down those dishes into raw ingredient data, determining the general healthiness of the food we cook.
I live in Chicago, where you can buy a hotdog (mustard, tomatoes and pepper — no ketchup) on every street corner and that molten cheese-and-sausage coronary known as the Chicago-style pie is always a phonecall away. The Windy City has a reputation for unhealthy eaters, but a new study by recipe search engine Yummly indicates Chicago isn’t the unhealthiest city in the country when it comes to cooking our own food.
That distinction goes to St. Louis.
Yummly, one of the largest recipe search sites on the web, analyzed over a billion individual data points from 100 million user searches over the last year to create a geographic breakdown of they types of recipes we Americans suss out online. Then it used its own recipe analytics to break down those dishes into raw ingredient data, determining the general healthiness of the food we cook (or at least the healthiness of the recipes we search for).
Yummly's recipe search app on the iPad (source: Yummly)
Yummly’s recipe search app on the iPad (source: Yummly)
The resulting trends may not surprise you. On the more “cosmopolitan” coasts, people tended toward healthier dishes, favoring fowl and fish over four-legged beasts. San Francisco, which scored highest on Yummly’s health index, is 30 percent more likely to search for a vegetable dish than Cleveland, Ohio, which scored third lowest.
We in the middle of this great land, well, we love our bacon, cheese and cake. And frankly I’m a bit disappointed that we Chicagoans — given our obsession with meat and fat stuffed into intestines — didn’t even rank in the top five of Yummly’s unhealthy index of the 25 most populous metro areas. In case you were wondering, though, here are the cities that did:
Source: Yummly
Source: Yummly
Fellow inlanders (and North Carolinians), it should make you feel better to know that according to Yummly’s data, California reinforced all of the expected stereotypes for kale and tofu consumption. So, we may die sooner, but hey, at least it will be with a belly full of meat.

0 comments:

Post a Comment