mark automatic import invalid

JimHokanson
I recently imported a paper and the automatic metadata detection was invalid. Does it make sense to flag that somehow as invalid? I don't know what system is used for automatic detection but if there is some learned rule, would allowing users to mark it as invalid be helpful in unlearning the rule?

Here's the paper:
https://www.nature.com/articles/s41592-021-01302-4

For me this detects the following paper and thus adds the incorrect metadata:
https://www.nature.com/articles/s41592-021-01314-0

Also, a +1 request for being able to clear metadata and reimport it for entries that were not recently imported but were determined to be incorrect. I know that request was made before but it is unclear if there has been any progress.

Finally, thanks for all the work on Zotero 7. I'm loving the new version!
  • adamsmith
    You're linking to the same paper twice.
    FWIW, it's pretty quick to just drag a PDF away from under the metadata and then manually create a new parent item. 3 clicks instead of one, but not exactly laborious, either.
  • JimHokanson
    Ah, thanks for the tip on the moving. At least for me dragging the PDF away changes the list of papers displayed so that the isolated PDF and the old metadata are not side by side. This means a bit of extra hunting to find the old metadata and delete it. Not that end of the world though.

    Regarding the same paper twice, I'm not following. Are you saying this is only happening on my account because of the papers I'm working with? The links above are different .... One of the papers above is on class imbalance and one is on an Australian biochemist. When I import the class imbalance one the metadata for the Australian biochemist is being pulled. Again, perhaps it doesn't make sense to have something that allows users to provide feedback on when the detection algorithm is working or not, but I figured I'd mention it. No worries if this won't be implemented any time soon.
  • adamsmith
    Sorry, I had just misread the links; ignore the 2nd part.
  • adamsmith
    Yeah, so this is wild -- the metadata for the incorrect paper is explicitly written into the metadata of the Imbalance paper. If you download it (with or without Zotero), open it in Acrobat, and look at the properties, here's what you see:
    https://s3.amazonaws.com/zotero.org/images/forums/u2433/ajh4cw8u2n0lxmc0l6lf.png

    Not sure there's much Zotero can do about that. It doesn't usually heavily rely on that data, but if there's an explicit DOI in there, it makes sense to use it.
Sign In or Register to comment.