SHARE 2.0 ranks records by a 0–100 score made entirely from five deposit-time buckets: Stewardship, Harmonization, Access, Reusability, and Engagement. Observed attention or downstream activity contributes no points. Evidence-support states in the published release ensure a field the source does not expose is not treated as demonstrated absence.
Ranking Algorithm Components:
- Five buckets, each worth 20 points; four points per passed signal.
- The denominator remains 25 even when repository evidence is unsupported.
- Every eligible record receives a terminal scored, unsupported, not-evaluable, or failed state in the release ledger.
Data Sources:
- Frozen normalized production records from each repository in the score release.
- Release and run identifiers bind results to source snapshot, scorer, adapter, and repository pledge versions.
- Native or file-level evidence not retained in the snapshot remains unsupported or not evaluable until a later enrichment release.
Algorithm Updates: The ranking formula is continuously refined based on community feedback. Major changes are documented in the changelog.