Visit the main Crossref website

Comment/tagging aggregation format? (e.g. pubpeer reviews)

Not a member, not a researcher. Just got interested in crossref for the neat API (crossref_commons works great). And due to a recent interest in e-cigarette junk science; whose proliferation is IMO partly caused by the medical journals and their lackluster UX.

Anyway, what I think (from my perspective) fairly interesting would be an aggregation system for study reviews. Not just studies referencing each other, but particular short assessments as pubpeer or sciencemediacenter did them. (Neither seems very active anymore?)
And since none of them have an API (or no official one), and some of these reviews are buried in the blogosphere/indieweb even; I thought about how a distributed comment collection format could be structured. TBH, didn’t give it much thought:

 {
     "http://0-doi.org.lib.rivier.edu/10/foobar2": {
          "tags": {
               "🔥": "smolder-study"
          },
          "comments": [
              {"note": "Detailed testing material/data absent"},
              {"url": "http://blog.post/xyz"}
          ]
     }
     "http://0-doi.org.lib.rivier.edu/10/xyz3": {
          "tags": {
               "🕥": "causal-transposition"
          },
     }
}

This would still require manually registering such pools. But OTOH allowed conversion of e.g. pubpeer into a simple JSON dump. Perhaps even a <link rel=crossref-collect href="/.well-known/crosscomments.json"> for detecting such archives.

Would something like this fall within crossref’s realm or interest? (Despite touchy topic and such).

1 Like

Thanks for this comment. We already do some of this through Event Data (see Event Data - Crossref). We don’t check PubPeer, but we’re currently looking at which new sources the community would like to see. We do already collect events from Reddit, Twitter, and Hypothesis which share comments and annotations.

2 Likes

Indeed. I should have looked at Events in more detail. (Assumed it was just about digitalised conventions/presentations.)

It’s not fully what I was hoping for. But then again, probably don’t need embedded excerpts in the API.
At the very least allows for building a bridge, if it covers Stackexchange and Reddit already.

Thanks! And my regards to the documentation team. Outstanding.

1 Like

Btw, Pubpeer does seem to have an API of sorts, at least for their browser extension:
https://github.com/PubPeerFoundation/PubPeerBrowserExtensions/blob/master/js/contentScript/pubpeer.js

It doesn’t quite follow the update feeds model, but individual doi: queries. So not sure if useful anyhow.

1 Like