Something with recipes

• 

  some of our amazing UI designs

Inspiration

M&S Beam Academy challenge

What it does

• adds a recipe's ingredients to a customer's cart
• suggests a recipe based on ingredients in the customer's cart
• search ingredients online on ocado
• scrapes a given web page containing a recipe for the ingredients
• text prediction when typing a recipe name

How we built it

• we used Java to parse JSON files containing recipe data, identify recipes with the most overlapping ingredients with the customer's cart, query ocado's database on items for ingredients, and scrape and parse online recipes for the ingredients they contain
• we used Python to build the recipe name text prediction using Good–Turing frequency estimation with pre-processed 3-grams
• we used Figma to design the UIs for all of our features to represent how we envision them working since we did not have enough time to implement a functioning version of the UI

Challenges we ran into

• sleep deprivation
• we didn't have enough time to implement a proper UI
• we didn't have access to any of the M&S APIs so we pretended we did
• the dataset for the recipe name text prediction was too big to finish processing before the heat death of the Universe so we reduced the size of the dataset (and therefore the accuracy of the model)

Accomplishments that we're proud of

• we're proud of the amount of things we achieved in the short amount of given time
• we're proud of the graphical representations of the UI that we made
• we're proud of the text prediction model

What we learned

• how to use Figma
• how to parse JSON data in Java
• sleep is optional
• how N-grams (e.g. the text prediction model) work

What's next for Something with recipes

• sleep
• inclusion of our features into the official M&S app

Built With

• figma
• github
• java
• json
• python