We merge the “business” table and the “review” table, and extract Chipotle’s information.
Yelp provides a comprehensive dataset in Yelp Challenge Dataset that includes all the merchants information such as addresses, tips, reviews, and reviewers. At last, Chipotle should improve the online ordering system and mobile app, in order to better solve the wait time problem and improve customers’ buying experience. The third one is to investigate the current quality of its most important product, burrito, which has gone fewer and fewer positive reviews.
#CHIPOTLE YELP HOW TO#
The second one is to investigate the supply chain to see how to improve the quality and taste of its proteins, especially chicken. The first one is to improve the freshness and quality of its ingredients to regain its strategic strength. Based on this, we can detect the pain points of Chipotle’s key business.īased on our analysis, we provide Chipotle with four suggestions in the areas of ingredient quality, wait times and digital ordering.
#CHIPOTLE YELP SERIES#
Do time series analysis of word sentiment, through which we see the monthly change of occurrence frequency of important words such as “line” and “burrito” in the positive or negative reviews.Draw a radar graph to show which topics are Chipotle’s advantages and which are its disadvantages. Use the LDA Mallet to do topic modeling to extract six topics from all reviews, and calculate the average sentiment scores for each topic.Draw word clouds for positive and negative reviews respectively to see what words appear frequently in positive or negative reviews.
The step-by-step guidance of python coding has been shared at Github: OverviewĬhipotle has been losing customers and isn’t performing well in the competitive fast-casual dining industry these years. This is the final project of our team ( Crystal Huang, Jack Ye, Hao Hao, Yuting Li) in Machine Learning class of UC Davis MSBA program.
0 kommentar(er)
