Sources & transparency
Editorial & corrections policy
FeelDB distinguishes source metadata, AI estimates and viewer responses so you can judge what a title’s emotional fingerprint is based on.
What sources do we use?
Catalogue metadata and identifiers come from TMDB and IMDb where available. Emotional estimates come from the Gemini model family using metadata and overview text; viewer ballots record selected emotions submitted by viewers. AI text and shares are estimates, not verified audience reactions, and neither source metadata nor ballots imply a professional editorial review.
How can you report an error?
Report errors through the existing contact form. Include the title URL, the field or statement that appears wrong and a supporting source when possible. This lets us compare the report with catalogue metadata or the relevant emotional data. Reporting an error does not guarantee a change or a response within a particular time.
Report an error via contactWhat does a correction or update mean?
A correction should reflect supported source information and preserve the distinction between AI estimates and viewer ballots. We do not convert an AI estimate into a viewer reaction. “Data updated” means the stored title record changed; it is not a claim that a reviewer checked the title. We do not promise a review schedule or a response deadline.