Community

As explained in Domain-Specific Small Language Models

A group of graph nodes that are more densely connected to each other than to the rest of the graph.

A community is a group of graph nodes that are more densely connected to one another than to the rest of the graph. In GraphRAG, communities are detected and summarized during indexing, then reused at query time to help answer questions over a knowledge graph. section 14.1

Used here to:

• cluster related entities before summarization. section 14.1
• support partial responses from multiple graph regions. section 14.1
• improve coverage and diversity compared with conventional RAG. section 14.1

GraphRAG workflow

Source text is chunked, entities and relationships are extracted into a knowledge graph, communities are detected, and each community is summarized for later query-time use. section 14.1

Figure 14.1 The logical blocks and workflow of Microsoft’s GraphRAG proposal

Shows the overall GraphRAG workflow where communities are created during indexing and consumed during query answering. section 14.1

How Domain-Specific Small Language Models covers it

How deep it goes

Defines the term and shows how GraphRAG detects, summarizes, and reuses communities, but stops short of full community-detection theory.

Examples

Uses a movie-database graph and the GraphRAG workflow diagram as teaching examples.

What it assumes

Assumes familiarity with graphs and knowledge-graph nodes and edges, and defines community from there.

What you can do after

Evaluate and apply community summaries to improve question answering over a knowledge graph.

How it connects

Connects to GraphRAG, knowledge graphs, entity extraction, modularity, and the Leiden algorithm.

liveBookmanning.com

Read Domain-Specific Small Language Models on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Graph Algorithms for Data ScienceCommunity structure is the organization of a graph into groups of nodes with comparatively dense internal connections.
• 

  liveBookmanning.com
  Financial AI in PracticeA group of graph nodes more densely connected to one another than to the rest of the graph.
• 

  liveBookmanning.com
  Essential GraphRAGA graph community is a group of graph entities that is more densely connected internally than to the rest of the graph.

Compare how 4 books explain community →

Related concepts

• graph densitymeasured by · Graph density is the proportion of possible edges between a graph’s nodes that are present.
• PageRankused by · PageRank is a link-analysis algorithm that estimates the relative importance of nodes in a directed graph from the structure and importance of incoming links.
• graph analyticsused by · Graph analytics is the analysis of entities and their relationships represented as a graph to identify patterns, properties, or predictive signals in its structure.
• community summarizationused by · Community summarization is the production of concise summaries describing groups of closely related entities, content, and relationships in a knowledge graph.
• label propagationproduced by · Label propagation is a graph-learning technique that iteratively propagates node labels through connections to assign labels or identify communities.
• graph clustering methodproduced by · A graph clustering method groups vertices into clusters based on their connectivity, structural patterns, or related attributes.

This explainer was generated by Manning's knowledge graph from the text of Domain-Specific Small Language Models. The section links point to the parts of the book it draws on.