Yelp Business Graph Analysis with EdgeDB

This project uses EdgeDB to explore graph-relational modeling and querying on a Yelp business and review dataset from Chicago. It focuses on modeling relationships among users, businesses, and reviews to uncover insights about customer sentiment and local engagement patterns.

Developed as part of the MSDS 420 course at Northwestern University, this work demonstrates practical graph-based querying strategies and evaluates EdgeDB’s expressiveness compared to traditional relational and document databases.

⸻

Features

• Schema Design: Defines a normalized graph schema using Link and Multi-Link types in EdgeDB.
• Data Ingestion: Loads CSV data into EdgeDB, transforming flat rows into a connected object model.
• Graph Queries: Retrieves insights such as most-reviewed businesses, average ratings, and reviewer activity.
• Comparison Discussion: Reflects on the benefits of EdgeDB’s graph-relational structure over traditional SQL and NoSQL models.

⸻

Key Insight

Graph-relational databases like EdgeDB blend the structure of SQL with the expressiveness of graph queries, enabling deeper insights from naturally connected datasets such as customer reviews.

View the source code on GitHub