Evaluation of Chunking and Embedding Strategies for Local Document Retrieval Using an Open-Source LLM in a Hospital
Chunking and embedding choices significantly affect retrieval in domain-specific RAG systems. While both Jinaai-v3 and Aari1995 were effective, they differed in stability, accuracy, and efficiency. Findings support deploying a locally executable RAG system for administrative use, guiding future optimization of chunking and parameter robustness.
Authors
- Jan Kirchhoff
- Fabian Berns (ORCID: https://orcid.org/0000-0002-7033-3789)
- Christoph Demus (ORCID: https://orcid.org/0000-0003-4319-3175)
- Jan Christoph (ORCID: https://orcid.org/0000-0003-4369-3591)
- Christian Jäger (ORCID: https://orcid.org/0000-0002-2740-8386)
- Jan Bossenz (ORCID: https://orcid.org/0009-0001-7946-8559)
- Annemarie Bianka Weise (ORCID: https://orcid.org/0009-0003-6925-5941)
- Carlo Günzl
Institutions
- Martin Luther University Halle-Wittenberg (DE)
Publication Details
- Journal
- Studies in health technology and informatics
- Published
- 2025-09-03
- DOI
- https://doi.org/10.3233/shti251383
- Primary Topic
- Biomedical Text Mining and Ontologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00