Skip to content

Add paper alignment runner

Marcel Konrad requested to merge paper-alignment into master

I initially created this runner to fix some modeling issues regarding papers, that are present in the production database. There, the metadata that is attached to papers via statements does not always follow the template of a paper. Sometimes, the values are stored using a literal, and sometimes they are stored using a resource. This was initially causing issues for !716 (merged). But since users do not have to obey the template structure when using the API directly, I ultimately changed my implementation for !716 (merged) to handle those scenarios as well. However, the alignment code still exists, and could be used if wanted. Although, I'm not too sure if we even want to merge it.

Some numbers:
As of 30.11.23, there are:

  • 72 papers whose publication month modeled using a resource
  • 102 papers whose publication year modeled using a resource
  • 1 paper whose url modeled using a resource
  • 1487 papers whose doi is empty (this did not affect !716 (merged) to begin with)

The resource/literal issue only affects papers older than 2.5 years. The doi issue affects papers older than july last year.

Edited by Marcel Konrad

Merge request reports