A Look at New York City Restaurants

Manasa Godkhindi
Posted on Dec 15, 2017

Introduction

Social data provides important, real-time insights on consumer opinion – on lifestyle, habits, brands, and preferences. Because these opinions are unsolicited, they provide genuine insight into consumer feelings, and, as such, they should be valued. Zomato, a research search and discovery website provides  information and reviews on restaurants. This blog provides an insights on the restaurant ratings and their services in Manhattan neighborhood.

I obtained data scraping data from Zomato. Data from dynamic web pages were fetched using Selenium. Data cleansing was done in python. Visualization and analysis including sentiment analysis was done using R. Analysis was portrayed using an interactive app built in shiny.

Research Questions

1. Does location have an impact on the pricing of the food?
2. Impact of rating based on reviews.
3. Customer opinion based on the reviews.
4. Emotions of customers towards restaurants.
5. What are the major concerns of customers?

Insights

Analysis - Cost, Rating, Location

Overall rating of restaurants are binned into 3 categories of excellent,good and average. Below plot is the representation of top 30 restaurants in Manhattan neighborhood based on their ratings.

Restaurants neighborhood do  have a say in the cost of the restaurants.Midtown and Gramecy-Flatiron neighborhood have high priced restaurants. Midtown is heart of abundant activity and a tourist heaven which justifies the high pricing.

Based on the analysis, Michelin starred restaurants and restaurants with added features like full bar,  private lounge also attributed to the high costs.

Sentiment Analysis

Sentiment Analysis, also known as opinion mining, is the process of determining whether a text unit is positive or negative. It can have a wide range of applications such as automatically detecting feedback towards products, news and characters or improving customers’ relation model.

To further analyze the sentiments of the customers, sentiment analysis was performed on the review.Words in each review were separated and the punctuation were removed so that a “bag of words” was generated for each review. Finally, we stemmed and filtered out the stop words.

Reviews were scored for different emotions like happy, sad, anger, frustration and also the polarity(positive and negative).

From the negative word lists, it can be observed that wait and crowd is really a taboo when it comes to restaurants. In addition, we can also conclude that restaurants should improve on their hospitality.

Friendly is one of the word that stood out in most of the positive reviews indicating the service of restaurants seems to be the priority for most customers

Below word clouds provides insights on the most commonly used terms in the positive and the negative reviews.

 

An interactive app  of the breakdown by cuisines and restaurants is available in my Shiny App.

Conclusion

Based on the analysis, metrics can be used by restaurants in improving hospitality, competitive pricing. Also restaurants considering opening in a specific neighborhood could use these insights to help them differentiate themselves from competitors.

A deeper study can be done using Aspect based analysis on the reviews to gather more information.

 

 

About Author

Related Articles

Leave a Comment

Your email address will not be published. Required fields are marked *

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags