[Crossmark] Different semantics for updated date between 2 major publishers

[gcabanac]

This post relates to Crossmark.

I am reporting a difference in the way 2 major publishers pushed metadata that are available through the REST API `updates` filter (see screenshots).

Let me show the issue based on 2 retracted articles.

Case 1: Elsevier https://doi.org/10.1016/j.heliyon.2024.e28357 (published 2024-04-15, retracted 2026-02)

The updated date field shows 2024-04-15 which is the publication date:

Case 2: Springer Nature https://doi.org/10.1038/ni829 (published 2002-08-05, retracted 2026-02-09)

The updated date field shows 2026-02-09 which is the retraction date:

Problem: The semantics of the updated date differs between the 2 cases.

I would expect to find the date of the update, just like in Case 2, isn’t it?

Could Crossref staff confirm this and ask the other publisher to update their records, please?

[ifarley]

Hi @gcabanac ,

Thanks for highlighting the discrepancies here. I am in contact with the technical contact at Elsevier. It looks to us like they may have deposited an incorrect date in the metadata. I’ll let you know what I learn.

Warm regards,

Isaac