Source Attribution
Source attribution identifies the specific source(s) that support a data point and the evidence type behind it.
Allocator relevance: Attribution prevents “AI-bluff”—it shows users what is known vs inferred and why they should trust it.
Expanded Definition
Attribution is the bridge between a field and its evidence. In allocator datasets, the same fact can appear differently across sources. Source attribution clarifies which source is used, whether the evidence is direct (explicit statement) or indirect (derived signal), and what date the evidence reflects.
Attribution is different from “source confidence”: attribution names the source; confidence expresses how strong the evidence is.
Decision Authority & Governance
Governance defines allowed sources, evidence hierarchy, and conflict resolution. When sources disagree, governance should preserve both values with confidence + lineage rather than overwriting silently.
Common Misconceptions
- If a field exists, it must have a reliable source.
- Attribution is only needed for legal/ownership fields.
- One source is enough for dynamic fields like roles.
Key Takeaways
- Attribution is how you earn trust at scale.
- Separate direct evidence from inferred signals.
- Handle conflicts transparently.