Look up for metadata from DOI for new item in library

Hello,
I moved to Zotero from Mendeley and really enjoy it, however, I am missing one feature.

If I add a PDF that does not have meta data, I can lookup for them in Mendeley by pressing the icon of magnifier close to the DOI or ISBN field and if only this field is filled, it properly fills the fields with correct data. Do I miss it or Zotero does not have this feature? This let's you add a new item very quickly knowing just the doi.
  • If you have a DOI, you can just add an item without ever having the PDF attache by clicking on the "Add by Identifier" ("Magic wand") button and pasting the DOI (or a list of DOIs for multiple items.

    If you have a PDF for which Zotero doesn't find metadata, you can right-click--> Create parent item, and Zotero will let you input a DOI or other identifier in the next screen.

    What Zotero currently doesn't allow you to do is to update/complete metadata for an existing item via DOI (or otherwise). That's planned and should not be too far out, but no specific ETA.
  • Thank you! So, if I add PDF for which Zotero finds wrong metadata, I am in the situation you described as third. Therefore, I have to use magic wand (found it!), drag'n'drop the PDF and delete the parent item with wrong metadata. This was the situation I was in. Thanks!
  • I would love to have a DOI lookup feature in Zotero.
    I have many entries which do not have the correct DOI metadata or wrong metadata.
    The current steps available in Zotero to fix the metadata are still much longer than what a simple DOI lookup would do.

    I understand that the main reasons for Zotero not to implement this feature are:
    1) The recommended way of adding entries to Zotero will most likely produce higher quality metadata. So people having this issue most likely aren't following the Zotero recommendations to import items.
    2) It is already possible to add a new entry with the DOI lookup, and then merge items to produce the same result. But this is still much longer than DOI lookup from an item with bad metadata.
    3) The most difficult step for developers to implement this feature is to decide what to do with the old metadata. I appreciate that the developers want to be extra careful not to delete any useful metadata. But for me, any item on which I would do DOI lookup has rubbish metadata anyway. So I do actually want to remove all the old metadata, without keeping any old metadata.

    I understand the reluctance of Zotero to add such DOI lookup. But:
    1) There are still many valid situations in which DOI lookup would be useful: import from Mendeley, wrong metadata extracted when importing from pdf files, cases where standard import would not be able to detect properly the DOI metadata, ...
    2) The simple brutal DOI lookup wiping away all the previous metadata is really the intended behaviour for me. I understand the concerns that this might be dangerous for some users, who may complain about losing metadata by accident. But hiding the option in the preferences with a clear warning should probably give access to this brutal DOI lookup only to the users who are aware of the consequences.

    Considering that DOI lookup is already implemented in Zotero, getting this option directly from the items with wrong metadata is probably not difficult to implement?
    Combined with the DOI Manager plugin, this would allow me to clean up many entries very quickly in my library.

    I have considered doing the DOI lookup in Mendeley, and import it back to Zotero. But the import of my Mendeley library in Zotero is taking around one to two days, so it is not very practical. And I understand that the import tool is only designed for single import rather than continuous sync between Zotero and Mendeley.
  • @mjthoraval: adamsmith literally said above that metadata updating is coming — you don't need to make the case for it. If you want an update on something, please, for your sake and ours, skip the theorizing about why we haven't implemented it and just post a one-line comment asking if there's an update. Here, the answer would be the same as we've given in various threads: a version of this feature is already implemented, we'll roll it out as soon as possible, but it's waiting on some long-planned changes to the data model that we'd like to make before it gets run across millions of items.
  • " If you want an update on something, please, ... just post a one-line comment asking if there's an update"

    Is there any update on this @dstillman?
  • Nothing new on this since July, no
  • I wish Zotero could have this feature too. It is so convenient.
    Since Zotero also supports creating an entry by dragging the pdf in, it would be better if we can force referesh the metadata from doi after Zotero reads the doi information from the document.
    It does not make sense to only support the other way - which is adding by doi and then drag the pdf in.
  • edited January 18, 2023
    "It does not make sense to only support the other way - which is adding by doi and then drag the pdf in."
    This sounds a bit strong. This is not the only supported way to enter a reference, nor the most efficient one: the standard way would be through the browser connector, which should takes care of everything in one click https://www.zotero.org/support/adding_items_to_zotero

    But I agree that the metadata update mechanism will be convenient when it comes to life :-)
  • I am here to leave a comment and show that this feature is very much awaited. Basically, I checked the news of this feature every several weeks.

    I am the type of people who add the references to Zotero by dragging the pdf files. Doing this way makes perfect sense to me, because one would want to read the pdf first before deciding whether the paper is worth adding. I hope the function will roll out soon.
  • I am the type of people who add the references to Zotero by dragging the pdf files.
    @yyhhoi: If you add a PDF directly to Zotero, either by clicking the save button in your browser or by dragging, Zotero will automatically retrieve metadata or try to extract basic metadata. If you're not happy with the metadata and want to redo it using a specific DOI, right-click and choose Undo Retrieve Metadata, and then use Create Parent Item and enter the DOI.
  • Hi,
    I am new to Zotero and still figuring out how things work. I too was looking for DOI look up for metadata and stumbled upon this link.

    I am using Zotero on Mint 21. When I right-click on the item or the linked pdf, I do not get the option saying "Create Parent Item". Am I missing any package?

    Also, When I drag and drop a pdf into Zotero, by default it makes a copy of the document in the local zotero folder. Is there a way to stop this?

    Thank you.
  • An ability to retrieve metadata on DOI was the core feature in my Mendeley workflow. Particularly, it was very useful in working with chapters, where wrong automatic DOI were very common. Manually rechecking had been done the job.
  • edited October 1, 2023
    @rishiddh: "Create Parent Item" only appears on standalone attachments. If you already have a parent item and want new metadata from a DOI, just use Add Item by Identifier, select both items, right-click, and merge them.

    If Zotero creates a parent item that you're not happy with (e.g., with the limited data it could scrape from the PDF itself), as long as it's within the same session you can right-click on the item, select Undo Retrieve Metadata, and then create a parent item from a specific identifier yourself.

    @emirsale: To be clear, you've always been able to create items by DOI in Zotero. What Mendeley's feature did was let you wipe out an item's existing metadata with new (often messy, incorrect) metadata from an identifier. You can do a version of that (but with much better metadata) in Zotero by using Add Item by Identifier as I describe above.

    Metadata updating in Zotero that lets you merge old and new metadata has been implemented and is coming soon. It's just waiting on some other technical changes that need to happen first.
Sign In or Register to comment.