Chapter 24 Introduction

This dataset is a subset of Yelp’s businesses, reviews, and user data. It was originally put together for the Yelp Dataset Challenge which is a chance for students to conduct research or analysis on Yelp’s data and share their discoveries. In the dataset you will find information about businesses across 11 metropolitan areas in four countries.We will do a detailed EDA, Text Mining ,Network Analysis and Geospatial Analysis on this dataset.The dataset can be found in Kaggle

For a city we spot the most popular business and also provide a map of the city of Las vegas with the business identified as dots in the map. We have analysed Las Vegas , Toronto and Phoenix.For Phoenix city, we also do Word Cloud, detailed Sentiment Analysis and Topic Modelling.

For a business we do the following analysis

  • Word Cloud of the reviews of the business
  • Top Ten most common Words reviews of the business
  • Sentiment Analysis - Postive and Not So Postive Words of reviews
  • Calculate Sentiment for the reviews
  • Negative Reviews
  • Positive Reviews
  • Most Common Bigrams (a collection of Two words) in the review text
  • Relationship among words
  • Relationship of words with an important word in the review such as steak, crab, food
  • Topic Modelling of the reviews

The business that we are analysing are

  • Mon Ami Gabi , a Las Vegas Restaurant , the most popular and highly rated restaurants
  • Bacchanal Buffet , the Second most popular and highly rated Las Vegas Restaurant
  • Pai Northern Thai Kitchen , the most popular Toronto restaurant
  • Chipotle Business in Yonge Street Toronto You guessed it right I like Chipotle :-)

How Sentiment Analysis can help your business ?

For a business, the Sentiment Analysis is very important. If the business owners can just see the Top Ten negative reviews, they can easily find out which aspect of the business they need to improve.

For example, the Pai Northern Thai Kitchen, the complaints were about Service. Now when we go deeper into the Service complaints, we can find out various aspects of the service complaints such as

  • why our waitress seemed to be in such a hurry to get us out of the place.

  • This restaurant was crowded and noisy. The tables were packed so closely that I was falling over other diners while maneuvering to my seat

  • but their service was God-awful. They rarely attended our table, It took 55 minutes for our food to arrive. They took our drink orders and did not deliver them

Another interesting complaint for the Chipotle Business in Yonge Street Toronto was that they did not accept Interac , a standard payment method in Canada

Examples involving it are as follows

  • Not complying with customers' choice to pay with Interac, a standard payment method in Canada, is also a nuisance

  • Only reason it got a 4 star is because they don't accept interac which is my go to.

How Topic Modelling can help understand your business and city ?

Topic modelling helps to pick specific topics from the huge volume of text. Topic Modelling on the Three popular restaurants and also on Phoenix City helps us to understand that complaints regarding restaurants and business is around Service