You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While looking into the data, I discovered an inconsistency in the calculation of amounts. The culprit are noticesUri that appear in more than one named-graph.
For instance, consider the notice with publication number: 00642806-2024.
Multiple publication numbers are associated with the same noticeUri: query
And this results in conflicting estimated amounts for the same noticeUri: query
A problem is that a notice URI has different publication numbers simultaneously. query.
We need to revise the URI pattern for a notice and/or named-graph of a notice. Using the uuid for the named-graph might be correct if only the last version of the notice is to be maintained in the triplestore, if all versions are to be maintained, then the named-graph should be based on the publication number. There are tradeoffs to analyze.
Some options from the top of my head:
URI of the named graph of the notice
Use the notice uuid, so one can keep only the latest versions of notices, by rewriting the named graph.
Use the publication number, so one can keep all versions of notices, (some queries become complex)
URI of the notice
Use the uuid, the same as the named-graph. Only one version exists
Use the publication number, the same as the named-graph
While looking into the data, I discovered an inconsistency in the calculation of amounts. The culprit are noticesUri that appear in more than one named-graph.
For instance, consider the notice with publication number: 00642806-2024.
Multiple publication numbers are associated with the same noticeUri: query
And this results in conflicting estimated amounts for the same noticeUri: query
A problem is that a notice URI has different publication numbers simultaneously. query.
We need to revise the URI pattern for a notice and/or named-graph of a notice. Using the uuid for the named-graph might be correct if only the last version of the notice is to be maintained in the triplestore, if all versions are to be maintained, then the named-graph should be based on the publication number. There are tradeoffs to analyze.
Some options from the top of my head:
Related: #41
The text was updated successfully, but these errors were encountered: