Is there any way to extract keywords from articles?

gonarias
Hello! First time posting here. I'm new to Zotero, and i was wondering if there was a way to extract keywords from articles and using them as labels for future work. I was handed a pretty big base of references, and as they come from EndNote, i only have the files with their titles, abstract, authors, year, etc. But most don't have labers (some do, i don't know why). Is there anything i can do to get keywords from these articles automatically instead of doing it manually?
  • DWL-SDCA
    edited July 6, 2023
    You said you are new to Zotero. Please say more so that someone may better help you...

    Did you import the articles to Zotero from EndNote? Do the records when still in EndNote have keywords? Many publishers don't provide keywords with their metadata. For the articles in Zotero do you have the Tag Viewer open or have you looked under the Tags tab? When items with keywords are imported from publisher sites or from some databases the keywords are placed into the Tags field.

    Info | Notes | TAGS | Related

    edited to add:
    If you import the following item from PubMed: PMID 37397707

    Do you see tags in your Zotero record? You should.
  • DWL-SDCA
    edited July 6, 2023
    I apologize if my post above was too basic for your needs.

    Are you asking if Zotero can automatically scan the title and abstract text of your existing records with tags and, from the text-words in each record, automatically add keywords/tags to new (or selected) Zotero records that have no tags? To my knowledge Zotero doesn't have this feature nor has it been requested before. Automating the creation of tags in new records based upon the tags and text in existing records might be appealing to some but would require a lot of computing power.

    Are you asking if Zotero can via the article DOI revisit the websites that hold the original article and capture keywords from the publisher website or database. Not yet.
  • gonarias
    Hello! Thanks for answering. I was wondering if some archives had metadata of their keywords and maybe Zotero was not retrieving them. I was thinking that because most of the files don't have tags (some do, so I guess it depends on whether the publisher makes them available). Maybe I was daydreaming because I have a pretty large database, and creating tags for each article would be very time-consuming, but it seems there is no easy way to get around that. I did some research, and maybe if I can find a program that scans the text and generates a word cluster, I could have a proxy of what the paper is about.
  • adamsmith
    Good chance this might happen in the future, it's just not currently available (it might be with one of the newer add-ons, but I don't *think* so)
  • dbrear
    Hi! It's a bit late to be answering your query, but if it's of any interest - I was spending too much time reading pdfs and tagging them, so a few weeks ago I got Claude to help me produce 'Tag / PDF Comparison tool.exe'. It produces a list of tags, taken from a standard text list, which match words found in the article. The tags are sorted by frequency into groups of ten so you can have an immediate idea of numbers.

    If this should be of interest, drop me a line at debbrear@gmail.com. It's about 23 Mb, so not emailable.
Sign In or Register to comment.