Incremental literature collection: workflow and methodological support by Zotero

Guest3713 · November 3, 2023

Hello, friends here. I would like to ask a question regarding the particular way you use Zotero to collect literature for a review. I know that Zotero has mutiple features, catering for different ways and processes throughtout one's research area. One may browse floods of articles to access cutting-edge techniques or concepts through Feeds with a few pieces added into the library from time to time, or may search articles and get .ris files from databases like Scopus and import and change keywords for another search and get another .ris file to import and on. Given that every search process would typically contain a few of articles that have been included in the existing library, I am trying to figure out the best way to include new items merely while not messing up with the old ones (notes, metadata, formats and so on) with the newly-searched duplicate records, which process I would name here as incremental literature collection. Thanks to this forum, which attracts not only software users but also researchers of different areas, I would like to seek advice for a detailed workflow for the mentioned mission, and to see if new features or improvement are required and can be find out to facilitate that job. Many thanks!

tim820 · November 3, 2023

As good as Zotero is, the ways in which you can check that you don't already have an item you are about to add is one of the less-than-efficient processes in Zotero. Mostly I find myself doing a manual lookup in my library ... a lot. Finding duplicates some time after addition when you check under Duplicate Items is not ideal; because you may not check it regularly, and so could add things like tags to a new item, only to find you have already done so with the same, previously-added item; or worse, you could mistakenly annotate two different copies of the same PDF.

So probably the simplest way to avoid this is just to check your Duplicates Items list very often. And merge the duplicates it finds, choosing the one with the best metadata. Annotations belong firstly to PDFs (and will apparently be visible in the main pane in a future version). In my experience, both copies of duplicate PDFs (or other attachments) have generally been kept by Zotero (under the single item kept) when duplicate items are merged (this may only apply to linked PDFs ?). So you have to explicitly delete the PDF you don't want to keep if you have two copies (ideally before merging; because once you do the merge, the item/PDF no longer appears in the Duplicates List).
https://www.zotero.org/support/duplicate_detection

Especially when doing bulk imports, it would be nice to be able to open the Duplicate Items list in a separate window, so duplicates would be seen immediately. It would also help identify which PDF to keep if PDFs with annotations had a different icon to those without annotations.

There is now a new plugin that attempts to detect duplicates at the time they are added (as well as offering some other metadata cleanup operations), but I haven't tried it. I am not sure if it works with RIS imports.
https://github.com/northword/zotero-format-metadata
"Note: The current version is only compatible with Zotero 7. If you are using Zotero 6, please download version 0.4.4 ." (the version for Zotero 7 has now advanced to v1.6.10)

Guest3713 · December 4, 2023

Many thanks to your nicely meticulous reply! I sometimes wonder if ChatGPT like machine learning models could be useful in helping deciding if two imported items are duplicates or not, but I guess the most economical way up to now is manual check anyway.