Sample dataset

Graph Intelligence includes a sample dataset that can be loaded into an existing Lakehouse for you to build a graph with. The blog Scoring Recommendations and Writing Back to OneLake uses the sample dataset to demonstrate the product’s capabilities.

  1. Open your Fabric workspace.

  2. Select New Item → Neo4j Graph Dataset.

  3. Select Sample retail dataset.

  4. Either create a new Lakehouse or select from an available Lakehouse and Connect.

New Lakehouses are created with schemas enabled.

  1. Create and wait for the tables to be created.

  2. Select Continue setup.

  3. Select from the available Organization, and Project(s).

  4. Select the region, required Memory, CPU, and Storage, and then Create.

  5. Copy and close the credential information, and store it in a password manager.

  6. Select all the tables, and Next.

  7. Review the AI-assisted graph model, make any necessary edits, and Transform to graph.

  8. Once OneLake tables have been transformed into a graph, you can use the Bloom and Query tools.