How QABO Matches AuthorsHow QABO Matches Authors

What

What rules does QABO follow, when matching Persons and Organisations to the Research Outputs coming into Pure? 

How

The author matching strategy is as follows: 

1. For each author the following information is provided by the profiling team
   1. Is the author an internal person in Pure, yes/no?
   2. If the author is an internal person in Pure, was the publication written while the person was at the current university, yes/no? I.e. whether the publication is internal for that author.
2. If the author is an internal profiled person, and the publication is internal to that author, the following is performed.
   1. The correct person is set, based on the information from the profiling team (using person ID).
   2. We then apply* organisations based on the authors organisation associations using the following logic
      • Organisations that were active at the published date of the given publication are added unless there is a primary organisation amongst them, then only that organisation is added.
      • If no organisations are found in 1) then active organisations are added unless there is a primary organisation amongst them, then only that organisation is added.
      • If no organisations are found in 2) then the first primary organisation or just the first organisation is added.
   3. An internal person in Pure will therefore always be associated to an internal organisation on the publication, if the publication is internal for that author as indicated by QABO (see 1.2 (yes) above)
3. If the author is an internal profiled person, where the publication is external to the author, i.e. it was not written while the author was employed at the current university (1b. (no) above).
   1. The correct person is set, based on the information from the profiling team.
   2. Scopus typically has information about what organisation it was written in context of, both the department and university. Here we use the university, only.
      • We match external organisations in Pure for matches based on scopus affiliation id or name. We do NOT use any fuzzy search, we require the name, address, id etc. to be unique matches to be considered as the same external organisation. I.e. if there is more than one organisation in Scopus for the same university there will likely also be duplicate external organisations in Pure, except if all the data is exactly the same.
        • If we match we use the matched external organisation, 
        • If we do not match we create a new external organisation (incl. scopus affiliation id).
   3. An internal person in Pure will therefore always be associated to an external organisation on the publication, if the publication is external for that author
4. If the author is an internal non-profiled person
   1. We then apply* organisations based on the authors organisation associations using the following logic
      • Organisations that were active at the published date of the given publication are added unless there is a primary organisation amongst them, then only that organisation is added.
      • If no organisations are found in 1) then active organisations are added unless there is a primary organisation amongst them, then only that organisation is added.
      • If no organisations are found in 2) then the first primary organisation or just the first organisation is added
      • if no organisations are found we add external organisations instead based on the Scopus data.
5. If the author is not an internal person in Pure, we associate an external person
   1. We match external persons in Pure with that name and Scopus author ID in Pure. We do NOT use any fuzzy search, we require the ID to be present if it is present in Scopus, otherwise the name must match exactly to an existing external person in Pure. I.e. if there is more than one profile in Scopus for that author in Scopus, there will also be duplicate external persons in Pure, except if all the data is exactly the same.
      • If we match we use the matched external person
      • If we do not match we create a new external person with that information. 
   2. Scopus typically has information about what organisation it was written in context of, both the department and university. Here we use the university, only, and perform the same logic as in 3b, where we associate an external organisation.

*Meaning Pure will disregard the organisations information on the record in Scopus and derive the organisations from your organisation associations on the matched person.
It will do this based on the persons organisation associations periods and the publication date, whether or not the organisations are active or marked as primary.