Could You Make a Public Data License Index?
From the earlier reference collection. Read the current software blog.
A public-data license index records each source’s stated reuse terms, the page supporting them, and the date those terms were checked.
Record claims with their source
A license index helps users find reuse information without pretending to make a legal determination. For every dataset, record the publisher, dataset name, the exact license label the publisher states, and a direct link to the authoritative page where that statement appears. Keep the page itself as the evidence; a secondary directory or a search snippet is not a substitute for the publisher’s terms.
If the publisher provides no explicit license, say that the terms were not found in the reviewed material. Do not infer permission from public availability, a downloadable file, or a government logo. The index should distinguish a named license from a terms-of-use page, a public-domain statement, and an unresolved status.
Model what users need
A compact record can include source name, dataset title, license or terms label, source URL, date checked, and notes about scope. Scope matters because one portal may host datasets with different terms, or a license may apply only to a particular file. Link to the page that covers the specific dataset wherever possible.
For example, if a portal labels one table as available under a named open license but says nothing about its attached images, the index should not extend the table’s terms to those images. Phrase the entry narrowly: “The dataset page states…” and identify the asset covered. Avoid summarizing complex conditions in a way that drops attribution or share-alike requirements.
Keep the index maintainable
Treat each entry as a record that can be rechecked. Store a last-reviewed date and a review status such as current, changed, or link unavailable. A scheduled link check can detect a dead page, but it cannot confirm that the legal text has not changed. Revisit important entries periodically and compare the page’s current statement with the saved label.
If the page changes, retain a short change note and update the visible entry. Do not overwrite history in a way that makes an earlier citation disappear; an audit trail helps explain why an older project relied on different terms. Avoid scraping or reproducing whole license texts when a descriptive label and authoritative link are enough.
Present uncertainty plainly
Use filters for publisher, subject, and status, and make the source link easy to open. Include a note explaining that the index reports what the linked source states and that users should review the applicable terms for their intended use. This is a description of the index’s scope, not a substitute for the source’s own conditions.
Test entries with a clear named license, a custom terms page, missing terms, a moved URL, and conflicting statements across pages. The interface should make unresolved cases visible rather than burying them. A useful index saves time by organizing evidence; it earns trust by separating that evidence from conclusions the source did not make.