Review Rover is a comprehensive Python-based tool designed to analyze vast amounts of customer reviews to uncover insights about business performance, particularly focusing on the reasons behind restaurant closures. By leveraging advanced natural language processing techniques, including BERT for topic modeling and sentiment analysis, Review Rover distills customer feedback to provide actionable intelligence for improving customer engagement strategies.
The objective of Review Rover is to analyze customer reviews from over 150,000 businesses, totaling approximately 6.99 million reviews. By applying BERT for both topic modeling and sentiment analysis, the project aims to identify the key factors contributing to restaurant closures. This intelligence is crucial for developing data-driven strategies to enhance customer engagement and improve business outcomes.
- Topic Modeling: Uncovers the most prevalent topics discussed in customer reviews.
- Sentiment Analysis: Classifies the sentiment (positive, negative, neutral) of each review.
- Restaurant Closure Analysis: Identifies the key reasons leading to restaurant closures.
The dataset used for this project is the Yelp Open Dataset, which includes:
- Reviews:Textual reviews of businesses
- Businesses:Information about the businesses reviewed
- Users:Data on the users who wrote the reviews
- Checks-in:Data on the check-ins of users at businesses
The analysis highlighted several key factors leading to restaurant closures, such as:
