/oneplacefoodie

One Place to Find Where to Eat Next!

Primary LanguageJupyter Notebook

One Place Foodie

One Place to Find Where to Eat Next!

Objective

  • To find a restaurant nearby filtered by location and cuisine.

Project Overview

  • Scraped over 27,000 Restaurant Pages from a food directory platform
  • Cleaned data and added geocodes
  • Plotted all restaurants on map box with hover-over tooltip

Code and Resources

Python Version: Python 3.8.5 Anaconda

Packages: pandas, numpy, matplotlib, seaborn, selenium, geopy, pickle, json, regex

Web Scraping

  • Food directory platform limited search results by page number. Suggestion was to use government food license titles to do a back search on the platform.
  • The above searching method resulted in over 300,000 website links. After removing duplicates, only 27,000 results were used.

Data Cleaning

  • Data scraped had many missing values and had to be cleaned with regex.

Production

  • Streamlit was used to deploy the dataframe visualization and filtering

alt text

Futher Improvements

  • Implement restaurant recommendation based on user's taste profile
  • Implement dataframe filtering on mobile app

Business Value

  • Collect data on food preference of users (anonymously)
  • Collect spending habits from restaurant visits
  • Opportunities for Restaurants to promote thier brand