Temporal GIS Narratives
Can sheaf/narrative structure preserve provenance and consistency in evolving geographic knowledge graphs better than event logs?
Hypothesis
A narrative/sheaf representation of changing spatial relationships stays more consistent under repeated edits than a plain event-sourced log, at a reasonable query-complexity cost.
Background
Time-Varying Data as Sheaves (arXiv 2609.09056) proposes a unifying formal language for time-varying, switching-relationship data — highly mathematical, no public code or benchmark yet.
Method
Simulate a parcel, infrastructure, or logistics-network domain with repeated changes; compare event-sourcing, temporal property graphs, and a narrative/sheaf representation.
Measurement Plan
Change consistency, provenance fidelity, and query complexity across representations.
Expected Artifact
An interactive temporal map plus a query benchmark, built on NetworkX/Neo4j and a mapping layer.
Limitations
The source paper is exploratory and mathematical — no empirical benchmark or public implementation exists to build on directly.
Status
This experiment is Proposed — queued, not started. Results, data, and measurements will appear here once it actually runs.
