Loading...
Loading...
Contributors:
Dataset URI:
10.5281/zenodo.23560100Supplementary study package for “Graph as Data, Graph as Organizer: An Evidence-Traceable Configuration Framework for GraphRAG”. This static two-configuration pilot compares SAP HANA Cloud RDF question answering using langchain-hana 1.2.0 with Microsoft GraphRAG v3.1.2 standard indexing and fixed-pool Global Search. The archive contains configuration manifests, the codebook, 56 primary coding records, 30 evidence locators, 12 pinned source snapshots with upstream licences, provenance verification records, revision events, the offline audit script and six negative controls, retained second-coder records, agreement comparisons, historical outputs, release notes, citation information, and SHA-256 checksums. Release v0.4.1 clarifies documentation for study v0.4 without changing the primary coding, codebook, evidence excerpts, or source bytes. Revalidation confirms 56 records, 12 source files, 30 resolving locators, 41 non-unknown records with resolving references, and 15 unknown records. Recomputed second-pass value agreement is 56/56. The retained metadata records AI-assisted coding using GPT-5.6 Sol and Claude Sonnet 5. The second pass shared working-session context with the original coding and lacks the separate source-line locators and excerpts required by its planned prompt. It supports recorded coding consistency, not independent blind replication or semantic correctness. No vendor pipeline, database query, or answer-generation endpoint was executed. See METHODS_CLARIFICATION.md for the evidential limitations. Original study data and documentation are licensed under CC BY 4.0; original audit scripts are licensed under MIT. Third-party source snapshots retain their upstream MIT and Apache 2.0 licences. See RIGHTS.md for the licence boundaries.
Keyword(s):
GraphRAG.
MeSH Term(s):
Grants and Funding:
No Grants or Funding Data Available
Associated Publication(s):
No Associated Publication Data Available
Supplementary Data:
No Supplementary Data Available
Related Dataset(s):
No Related Dataset Data Available
The National Library of Medicine (NLM) provides information about biomedical datasets. Inclusion on an NLM Web page does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: NLM Web Policies (nih.gov)