We propose a Graph Retrieval-Augmented Generation (GraphRAG) framework for the analysis of Semantic Verbal Fluency (SVF) tests, combining graph-structured representations of lexical retrieval with retrieval-conditioned generation. The approach preserves the relational and temporal structure of verbal production while grounding language-model outputs in verifiable graph evidence, addressing the limitations of both aggregate-score analyses and unconstrained generative interpretation in longitudinal cognitive assessment.
Graph-based analysis of verbal fluency tests for cognitive decline assessment
Ortu, Marco;Carta, Andrea
2026-01-01
Abstract
We propose a Graph Retrieval-Augmented Generation (GraphRAG) framework for the analysis of Semantic Verbal Fluency (SVF) tests, combining graph-structured representations of lexical retrieval with retrieval-conditioned generation. The approach preserves the relational and temporal structure of verbal production while grounding language-model outputs in verifiable graph evidence, addressing the limitations of both aggregate-score analyses and unconstrained generative interpretation in longitudinal cognitive assessment.
File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
iris_paper_GRAPHRAG.pdf
Solo gestori archivio
Dimensione
2.42 MB
Formato
Adobe PDF
|
2.42 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I metadati presenti in IRIS UNICA sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono protetti da diritto d'autore, salvo diversa indicazione.
