Skip to main content

Creating a Knowledge Graph

Once your dataset is ingested in Amplifi, you can generate a Knowledge Graph to extract and organize entities and their relationships.

Step 1: Start Graph Creation​

After ingestion, use the Add Graph to Dataset option to initiate graph generation. Amplifi will extract entities from your dataset using built-in models.

You can choose between:

  • Extract from Dataset: Let Amplifi automatically detect relevant entities.
  • Define Manually: Manually enter specific entities you want to include.

Step 2: Review and Select Entities​

Once entities are extracted, Amplifi shows a list of detected entity types. You can review and select which entity types to include in the graph.

Step 3: Generate the Graph​

After confirming the entity types, click to generate the graph. The system will process your selection in the background and create a structured network of entities and relationships.

Step 4: View Graph Summary​

Once the graph is ready, a summary is shown with:

  • A list of entity types
  • A count next to each type indicating how many entities were identified in that category

This helps you understand the distribution of entities across your dataset.

Step 5: Explore Relationships​

Switch to the Graph Relationships tab to see the relationships discovered between entities. These connections form the backbone of the Knowledge Graph and help link concepts together.

Step 6: Manage the Graph​

You can choose to:

  • Regenerate the graph if the data changes or you want to modify entity selection
  • Delete the graph if it’s no longer needed

Creating a Knowledge Graph helps you uncover meaningful relationships in your data, improving both semantic understanding and search relevance.