What the platform measures—and what it does not
Scholarly metadata
ResearchRank retrieves author, institution, publisher, work, citation, topic and identifier metadata from the configured scholarly API. Source identifiers are retained so profiles can be rechecked and probable duplicate records can be flagged.
Refresh and caching
API responses are cached for performance with shorter caching for search and longer caching for stable profile metadata. The ranking calculation itself is performed server-side from the current source response and declared filter context.
Entity disambiguation
Source IDs and ORCID, when present, are stronger identifiers than names. Similar names alone are never automatically merged. Probable duplicates are identified using ORCID, normalized name, institution overlap and output similarity and are presented for review.
Coverage limitations
Bibliographic databases can contain missing works, delayed citations, affiliation errors, merged or split author records and uneven field coverage. A mathematically exact calculation can still reflect imperfections in the underlying metadata.
Responsible use
ResearchRank analytics and certificates are supplementary evidence. Formal hiring, promotion, funding or institutional evaluation should consider qualitative peer review and multiple independent indicators.