An application which helps people find their foodie partner!
- Find a Foodie-Partner
- Search Restaurants based on Location and Rating
- Sign In and Sign up Page
- Profile Page: User can add more details about himself and can add his preferences of Cuisines and can also edit them later.
- Then the user have options to Find Match or to Find Restaurants based on their preferences.
- After going to find Match he can update cuisine and then will get 4 foodie-partners to choose from.
- When he goes to Find Restaurants he will have to enter the location and based on it he will get suggestions of top restaurants based on the rating found from Zomato API.
- Location personalization for recommendations.
- Re-filtering of the recommendations specifically for age and gender preferences.
PS: We maintained pickle files for the model as well as the dataset which would be used for the recommendations. as it is faster to load files from pickle rather than getting the same data from database. These pickle files are automatically updated whenever a new user add entries for cuisines+ratings or whenever they update ratings for the cuisines.
-
All the configurations are managed through the file
config.json. Changes in it reflect to the complete application. -
Note that there is a
db_config.jsonfile that we made for configuring with database and is not present on GitHub (Obviously!) so be sure to create one and edit accordingly.
- Fork the repo.
- Install dependencies using
requirements.txt. - Right now database is not pushed onto GitHub so consider creating a database using MySQL Workbench or any editor you prefer.
- After creating of database and installing deps, run
front-end/main.py. This will start a local server for flask and the application will be up for working.
For getting a preview of how the model was created or visualizations, please refer to BonHacketiet.ipynb.
Link to video submitted at hackathon - https://www.youtube.com/watch?v=i9S8DIqNA9E