A Human-in-the-Loop, LLM-Centered Architecture for Knowledge-Graph Question Answering

DOI10.48550/arXiv.2602.05512MaRDI QIDQ7266596OpenAlexarXivFDO

Authors Larissa Pusch, Alexandre Courtiol, Tim O. F. Conrad, Tim Conrad

Publication date 5 February 2026

Published in arXiv

Tldr This paper presents a human-in-the-loop architecture that combines large language models with knowledge-graph retrieval to answer complex questions in knowledge-intensive domains while mitigating hallucinations. The approach grounds model outputs in structured knowledge graph data, improving explainability and factual accuracy over text-only retrieval-augmented generation.

Full work available at URL https://doi.org/10.48550/arXiv.2602.05512

arXiv classification

cs.CLcs.IR

Cites work

• Bravo MaRDI: A Wikibase Powered Knowledge Graph on Mathematics
• Knowledge graphs
• Simultaneous inference in general parametric models
• Testing environmental and genetic effects in the presence of spatial autocorrelation

This page was built for publication: A Human-in-the-Loop, LLM-Centered Architecture for Knowledge-Graph Question Answering

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7266596)