This project explores the use of EdgeDB to model and analyze a real-world graph-relational dataset, Yelp businesses and reviews from the Chicago metro area. Built for the MSDS 420-3 Databases course at Northwestern University, the notebook walks through schema design, data loading, and analytical queries over a semi-structured dataset.

By adopting EdgeDB, the project illustrates how flexible, queryable schemas can unlock new dimensions of insight, blending the rigor of relational databases with the adaptability of graph systems. Business metadata, category tags, and review sentiment are cross-linked to enable multi-hop queries that mimic human curiosity.

Features

Key Insight

SQL-based tools dominate enterprise analytics, but graph-relational platforms like EdgeDB offer a more natural fit for richly connected data. This project shows how the right data model can turn flat tables into exploratory engines for user and business insight.

View the source code on GitHub

View on GitHub →