Data and ranking methodology
How YunoDB collects, filters, translates and ranks software data.
Public data sources
YunoDB imports structured public information from supported source adapters. Depending on the source, this can include descriptions, project activity, popularity signals, licenses, official links and update dates.
Quality filtering
Every item passes deterministic checks before publication. The quality score combines source quality, popularity, content completeness and project metadata. It is a discovery signal, not a product review or guarantee.
- Archived or missing projects are not automatically published.
- Duplicate source records are consolidated where reliable identifiers match.
- Thin records are excluded from indexable listings.
- Paid placement does not change the quality score.
Categories and comparisons
Categories are assigned from source metadata and maintained taxonomy rules. Alternative and comparison pages are published only when enough related, visible tools exist. Comparisons use factual fields available in YunoDB rather than invented feature claims.
Translations and corrections
Traditional Chinese summaries are generated from the source text with instructions to preserve brand names and facts. Automated translations can still contain errors. Corrections are welcome through the contact page.