Word Cloud

Above is word cloud of dishes which also appeared on menus with Eggs Benedict.

For the word cloud, we used a smaller sample set of menus with standardized dish title to find out what items had similar popularity to eggs benedict across different menus. The team found many dishes that centered around one type of meat “chicken”, “steak”, and many popular entries existed as supplemental to other breakfast items such as “fruit”, “cheese” and “coffee”. Some items likely more popular in a pre-1950s setting like “liver” and “kidney” appeared in multiple menus. Each menu only had about one instance of “eggs benedict” or “eggs florentine”, making it relatively small to other consolidated items, but puts it in roughly the same realm of popularity as “milk toast”, “bacon”, “toast”, “poached eggs’, and “ham and eggs".

The word cloud of similar was a team effort of standardizing menus to handle differences in the dish titles. While we first tried to compile all the menus OCR text to create a word cloud, the team ran into limits using OpenRefine to pull thousands of different Menu Items across hundreds of pages of menus out of the full dataset. The team instead used a smaller sample set of menus which ranged across different types of restaurants the team felt were representative of the menus. We then took over one thousand titles of dishes and cleaned up the text to remove punctuation, numbers, odd descriptions, and standardized different names of dishes (ie Eggs Benedict, Eggs Benedictine, Poached Eggs Benedict), then we could see patterns emerge among which related dishes were popular.