Data and learning
Preview only, nothing is sentThis page explains what GapDossier would learn from how the product is used, so the workflow can be improved. Nothing on this page collects or sends anything today, and nothing here changes your results.
Notice version 0.1-draft. Wording is a draft and is pending legal and privacy review.
What would be recorded
- Which step of the workflow you reached, and whether it finished or stopped
- The kind of identifier used, for example DOI or PMID, never the value
- How long a step took, grouped into broad ranges
- Which view or tab you opened, and whether an export was requested
- A safe error label when something did not work
What is never recorded
- Publication identifiers you enter, such as a DOI, PMID, PMCID, or OpenAlex Work ID
- Titles, abstracts, excerpts, notes, and any other research content
- Search text, uploaded files, or export contents
- Names, email addresses, and anything that identifies you or a colleague
Your research content stays separate
Product improvement and any use of research content for model training are separate things. Research content is never used for training unless you choose it, and that choice is currently off.