No reliable means of identifying authors

[Raza Ali]

We currently store the first name, last name and initials for authors. It is possible to have multiple people with identical names on publications, therefore it is necessary to create duplicate entries for what may be the same person on different papers. It would be good to have a unique identifier (e.g. orcid) when one is supported by the pubmed service.

[Camille Laibe]

I would vote for not duplicating the authors for the time being. When EuropePMC records all ORCID for all the publications' authors, I think that way it would be easier to refresh the database content and create the necessary new authors. But maybe the Issue #13 would impact on this decision.