ForkOut: Restaurant Metadata Extraction from Unstructured Menu

This project includes using helping ForkOut match restaurants to users based on their tastes using restaurant metadata.

Project Lead:

ForkOut is a mobile application that makes restaurant reccomendations based on users taste preferences.

Detailed Project Description:

ForkOut recommends restaurants to users of the app. Metadata about a restaurant is required to match a user’s taste with the restaurant. While Yelp has some metadata about a restaurant via their api, it is restricted in terms of what it provides. We want a robust and reliable way of categorizing restaurants based not only on cuisine type but on a particular dish as well. This project would help us match restaurants to users without relying on Yelp data and provide a unique search experience that doesn’t exist on the market currently.

Technical Components:

Provide assistance in the following topic areas: data analytics, data mining, software engineering, mobile application development, and machine learning. 

A restaurant coordination app needs data to be scraped from restaurant websites and extract menu items, cuisine type and price range from it. The list of restaurants can be compiled from google places api search. The results will typically include a website url for the restaurant. The menu could be discovered by crawling the site.

Data Set(s):

The datasets required for the project should be scraped from public online websites. Google has api to discover restaurant websites to scrap and restaurants

Skill/Expertise Requirement(s):

1) Web scraping
2) Parsing html (to extract relevant data)
2) Machine learning (to classify restaurants to a particular genre based on the menu items offered)