BigData - Food edition
A wacky paper in Nature, where they create a graph of recipes based on the ingredients used. The secret sauce (funny!) behind the graph is that the more shared flavor compounds in the nodes (ingredients), the closer the nodes are in the graph.
The result looks something like this
This is, of course, a highly simplified representation - do go to the original for many many details about how they pruned the graph, and how they validated that the graph was still accurate).
Some crazy details
North American food has cane molasses as one of its primary ingredients. I ask you, why? I mean, whats so special about cane molasses? I know it gets used as an ingredient in alcohol, but the authors don't really give context as to what they mean.
Garlic and scallions go together in East Asian cuisine. Again, this is something that makes sense ex post facto, but I never realized that it was this common
North American and Western European cuisine is similar. Ok, thats a gimme. The funny one is that Southern European, Latin American, and East Asian have commonality. And that garlic is the key common factor!
The result looks something like this
Some crazy details
Garlic and scallions go together in East Asian cuisine. Again, this is something that makes sense ex post facto, but I never realized that it was this common
North American and Western European cuisine is similar. Ok, thats a gimme. The funny one is that Southern European, Latin American, and East Asian have commonality. And that garlic is the key common factor!
Comments