Performance Summary

Hybrid GraphRAG

Vector

Graph

Comparision across all questions

0.000RAGAS Correctness

Hybrid Graph RAG outperforms pure Vector search by +12.0% in factual accuracy.

precisionvs Vector

0.000

0.0%

recallvs Vector

0.000

+15.0%

Scenario Analysis

3 of 3 methods

0.75

0.5

0.25

Hybrid GraphRAG: 0.560

Vector: 0.500

Graph: 0.400

correctness

Hybrid GraphRAG: 0.160

Vector: 0.160

Graph: 0.070

precision

Hybrid GraphRAG: 0.560

Vector: 0.410

Graph: 0.550

recall

Hybrid GraphRAG: 0.250

Vector: 0.220

Graph: 0.120

fMeasure

Method	Correctness	Precision	Recall
Hybrid GraphRAG	0.5600	0.1600	0.5600
Vector	0.5000	0.1600	0.4100
Graph	0.4000	0.0700	0.5500

* Precision, Recall, and F-Measure indicate lexical overlap. METEOR accounts for semantic similarity. Correctness evaluates factual accuracy.