I. Project scope

This project aims to generate top 5 restaurants in Las Vegas based on Yelp reviews. More specifically, we give reviews different weightings based on the reviewer's expertise in a particular cuisine. As a results, our score will be more "authentic" than the raw average score used by Yelp. For this particular project, we will only consider three different cuisine: Chinese, Indian, and Japanese.

II. Data & datasets

Data source: Yelp Dataset Challenge Round 6 (JSON) http://www.yelp.co.uk/dataset_challenge

Cleansed Datasets: i business_subset 18 KB ii review_subset 4.7 MB iii user_subset 17 MB

i. 	business_subset
	- Subset of the business data
	- Will link this data via unique ID to review
	- Filters:
		city:		Las Vegas
		categories:	'Chinese', 'Indian', 'Japanese'
		reviews:	> 100
		stars:		>= 4.0

ii. review_subset
	- Subset of the review data:
		Review records for all restaurants selected in i, regardless of the user
	- Reviews are linked with business by business ID and with user by user ID

iii.user_subset
	- Subset of the user data:
		Distinct users of all review records selected

III. Web framework / UI

React web page styled with bootstrap. See user manual on how to load the page using webpack. The UI has a simple drop down list to choose the cuisine in interest.