Build Communities

POST

Creates communities in the graph by analyzing entity relationships and similarities.

Communities are created through the following process:

  1. Analyzes entity relationships and metadata to build a similarity graph
  2. Applies advanced community detection algorithms (e.g. Leiden) to identify densely connected groups
  3. Creates hierarchical community structure with multiple granularity levels
  4. Generates natural language summaries and statistical insights for each community

The resulting communities can be used to:

  • Understand high-level graph structure and organization
  • Identify key entity groupings and their relationships
  • Navigate and explore the graph at different levels of detail
  • Generate insights about entity clusters and their characteristics

The community detection process is configurable through settings like:

  • Community detection algorithm parameters
  • Summary generation prompt

Path parameters

collection_idstringRequired

The unique identifier of the collection

Headers

X-API-KeystringRequired

Request

This endpoint expects an object.
graph_enrichment_settingsmap from strings to anyOptional

Settings for the graph enrichment process.

run_with_orchestrationbooleanOptional

Response

Successful Response

Errors

Built with