Database-Driven Yelp Analytics

This project uses EdgeDB to construct a graph-relational schema over Yelp’s business and review data, enabling advanced queries and structured analytics. Built for the MSDS 420-3 course, it highlights database design, data modeling, and query optimization.

Yelp’s raw data is transformed into a semantic knowledge graph that connects businesses, users, and reviews through typed relationships. From here, advanced queries reveal spatial patterns, user preferences, and categorical trends across Chicago’s business ecosystem.

⸻

Features

• Graph-Relational Modeling: Implements a hybrid data model using EdgeDB to connect relational tables with graph relationships.
• Normalized Schema: Structures Yelp data into entities like Business, User, and Review, with integrity constraints and indexes.
• Advanced Querying: Demonstrates queries using EdgeQL to extract insights like top-rated categories, prolific reviewers, and spatial outliers.
• Chicago-Focused Analytics: Filters the dataset to spotlight insights from Yelp businesses located in Chicago, IL.
• Visualization: Exports data to CSV for visualization of review patterns, business clusters, and user sentiment analysis.

⸻

Key Insight

Database-first AI products require intentional design. This project shows how structuring and enriching data upfront leads to faster, deeper insights, with less reliance on LLM guesswork.

View the source code on GitHub