document copyright extraction

[Aazhar]

Hello @kermitt2 ,

I've remarked that from the extracted TEI, the copyright statement found under availablity tag is actually the publisher, is there any reason for this :

https://github.com/kermitt2/grobid/blob/master/grobid-core/src/main/java/org/grobid/core/document/TEIFormatter.java#L175

[kermitt2]

Hi Achraf !

By default, the copyright holder is the publisher. This is not a bug, this is a correct legal fallback :)
We can put something else only if the publisher waives its default copyrights to another license like CC.

I am open to solutions on what and how to fill this space... if we match with CrossRef we can get some copyright/license information. Of course this is relevant to the published version, and it might not be relevant to a preprint version for instance that is processed. It's hard to detect.

Otherwise, Grobid does not explicitly extract license information. It has simply a quite loose field "copyright" and I've never really tried to evaluate it. I am not so sure where to put the copyright field as extracted by Grobid to be honest, either under <availability><p>...</p> or under a dedicated <note> for more safety.

[Aazhar]

besides the mention to the publisher the extracted copyright statement could go inside the licence tag, it seems suitable, what do you think ? : https://www.tei-c.org/release/doc/tei-p5-doc/fr/html/ref-licence.html

Indeed, the actual copyright extraction might need some improvements, but I think the markers are not that diverse ?