Crate oxirs_graphrag 

OxiRS GraphRAG

GraphRAG (Graph Retrieval-Augmented Generation) combines vector similarity search with graph topology traversal for enhanced knowledge retrieval.

Architecture

Query → Embed → Vector KNN + Keyword Search → Fusion → Graph Expansion → LLM Answer

Key Features

• Hybrid Retrieval: Vector similarity + BM25 keyword search
• Graph Expansion: SPARQL-based N-hop neighbor traversal
• Community Detection: Louvain algorithm for hierarchical summarization
• Context Building: Intelligent subgraph extraction for LLM context

Example

use oxirs_graphrag::{GraphRAGEngine, GraphRAGConfig};

let engine = GraphRAGEngine::new(config).await?;
let result = engine.query("What safety issues affect battery cells?").await?;
println!("Answer: {}", result.answer);

Re-exports

pub use config::GraphRAGConfig;

pub use graph::community::CommunityDetector;

pub use graph::traversal::GraphTraversal;

pub use query::planner::QueryPlanner;

pub use retrieval::fusion::FusionStrategy;

Modules

config

GraphRAG configuration

generation

Answer generation module

graph

Graph processing module

query

Query processing module

retrieval

Retrieval module for GraphRAG

sparql

SPARQL extension functions for GraphRAG

Structs

CommunitySummary

Community summary for hierarchical retrieval

GraphRAGEngine

Main GraphRAG engine

GraphRAGResult2

GraphRAG query result

QueryProvenance

Query provenance for attribution

ScoredEntity

Entity with relevance score

Triple

Triple representation for RDF data

Enums

GraphRAGError

GraphRAG error types

ScoreSource

Source of entity score

Traits

EmbeddingModelTrait

Trait for embedding model operations

LlmClientTrait

Trait for LLM client operations

SparqlEngineTrait

Trait for SPARQL engine operations

VectorIndexTrait

Trait for vector index operations

Type Aliases

GraphRAGResult