Module 2

Working with Virtual Graph on Databricks

In this module you will generate the fraud ring dataset in Databricks, connect Neo4j Virtual Graph to it and run Full definition for Cypher (opens in a new tab)Neo4j's implementation of GQL, the ISO standard query language for graph databases. It is declarative: you describe the pattern to find, and the database decides how to find it. queries to detect fraud rings.

You will learn how PySpark handles data generation and Delta table writes, how to map the schema manually in Virtual Graph and how to write fixed-hop Cypher ring detection queries against a Databricks backend - including the key difference in Full definition for property (opens in a new tab)A named value stored on a node or a relationship. name casing compared to Snowflake.

Notebooks

Download the notebooks from the course repository.

  • The dataset is generated by 01_dataset_generator_db notebook, which runs inside the warehouse.
  • The SQL baseline queries are demonstrated in the 02_sql_baseline_db notebook.
  • The Cypher cheat sheet and ground truth validation live in the 03_cypher_reference_db notebook.

These notebooks are provided to help you complete the labs.

Ready, let's go!