What question did this study set out to answer?

To evaluate 1,200 edges within knowledge graphs across multiple domains and evidence levels.

March 17, 2026Open Access

Stratified Evaluation of Evidence-Graded Knowledge Graph Edges: A 1200-Edge Multi-Domain Assessment

Key Points

To evaluate 1,200 edges within knowledge graphs across multiple domains and evidence levels.
Conducted a stratified evaluation across four domains and four levels of evidence.
Employed Cochran-Mantel-Haenszel analysis to assess domain effects.
Utilized Firth logistic regression to handle sparse data cells.
Reported results with Wilson confidence intervals for statistical interpretation.
Identified significant domain effects in knowledge graph edges.
Provided insights into the relationship between evidence grading and edge significance.
Enhanced understanding of multi-domain interactions within knowledge graphs.

Abstract

Paper 15 in the Prioris research series. Presents a stratified evaluation of 1,200 edges across four domains and four evidence levels. Reports Cochran-Mantel-Haenszel analysis of domain effects, Firth logistic regression for sparse cells, and Wilson confidence intervals. Part of the Evidence-Graded Knowledge Graphs research programme (Paper 0: 10.5281/zenodo.19024611).

Stratified Evaluation of Evidence-Graded Knowledge Graph Edges: A 1200-Edge Multi-Domain Assessment

Key Points

Abstract

Cite This Study