Captain Cook
Captain Cook is a project developed by an EPFL team to analyse a recipes database and visualise the informations in a cool way.
Here you can see the results of the project! Enjoy!
The DataSet was constituted of 47000 recipes coming from different websites.
Because of this, it was important to clean the Data before using it.
The recipes were found on websites like All Recipes, Eating Well or Yummly.
This meant we had lots of recipes, but as every website is different, we needed to clean the recipes to have a coherent structure for all recipes.
We chose to keep the following informations:
In the end, we have a total of 47000 recipes to analyse.
One of the first things we are interested in is discovering which regions have the most recipes published.
Unfortunately, most recipes were not labelized. The 15700 labelized recipes came from 36 different regions, which is still quite a lot: we could eat only italian plates for dinner every day for almost 7 years!
On the side, you can see the percentages of the recipes per region in a chart (a PIE chart... get it?). It is clear that the community with most recipes is the US, which is mainly due to the fact that the websites we scraped are all anglophones.
At the second place come Italian recipes, followed closely by the mexican. The first 10 regions make for the 45% of all recipes!
What are the words people use the most when writing the title for their recipes?
Titles need to be catchy to interest people: apparently, most users really like chicken! We can also see that one of the words that people use the most is 'easy': people want recipes that are easy and fast to make, like salads or soups. A lot of recipes are about sweets: cookies, cake and pie seem to enjoy a lot of attention.
Obviously all around the world people are crazy about salt and pepper! But not in the same proportions. In average, 10% of the ingredients in a recipe is salt, the Belgian recipes are the saltiest whereas the Asian recipes are the least with only 3% of occurences. Countries are defined by their culture and recipes, try to identify them by their recipes' ingredients!
Are you willing to try something exotic? Just look at this chord diagram, it will help you find which regions have the more similarities yet with some differences as you have seen right before. We can observe some obvious correlation in this chord map, for example US and Canada shares 10% of their recipes. But if you prefer Greek food you will surely like Italian or Australian recipes.
Now that you know what you are looking for, it might be useful to see if you have the time for it... By looking at the map we see a lot of disparity in the average cooking time. German recipes takes quite a long time to cook, with their delicious cakes and patisseries, whereas US recipes are quite quick to realize but don't be fooled by junk food!
Maybe it is time to avoid fast-food and discover which regions offers the best recipes for your health. Korean recipes seems appropriate for low-carbohydrates diet but beware of their sodium level, they are the worst! Maybe you should go for a Japanese recipe, it seems healthier but if you don't like asian food, you should try Mediteranean or Scandinavian recipes.
