What problem does it solve?
The Polder dataset's automated code resolver fails to match a subset of person proposals from Dutch public sector appointment feeds: these include proposals with no matching family name in existing records, proposals with common family names that have multiple candidate matches, and proposals missing required birth year data for record creation. Manual resolution of these edge cases is required to maintain dataset accuracy and completeness without introducing duplicate or incorrect personnel records.
Core Features & Use Cases
- No-Match Resolution: Cross-references Wikidata candidates for proposals with no existing family name matches, validating name alignment, plausible age based on appointment start date, and role/organization fit to create new valid personnel records.
- Ambiguous Family Disambiguation: Compares candidate records with matching family names against proposal given names/initials, existing organizational affiliations, and mandate timeline overlap to identify the correct existing record or flag for manual review.
- Birth Year Completion: Fills missing birth year data for partial match records by cross-referencing existing record slugs, candidate birth data, and external authoritative sources to meet the dataset's slug naming requirements.
- Use Case: When processing a new State Gazette appointment for a director at the Inspectorate of Education that does not match any existing Polder person record, this skill resolves the proposal by validating Wikidata hits, confirming role and start date alignment, and generating a valid new record entry with a confirmed birth year.
Quick Start
Use the lookup-person skill to resolve the unresolved person proposal for the new director at the Inspectorate of Education included in the latest State Gazette appointment data.
