What would be a good (easy, quick, replicable) way to get a random selection of, say, 1000 papers (DOIs) based on specific criteria, such as that they ought to have been:

  • published in journals of a specific discipline (e.g., Political Science)
  • published in a specific year (e.g., 2015)

I am agnostic about the data source. However, openly licensed sources are preferred over other ones. For instance, CrossRef or OpenAlex would be preferred over Scopus or Web of Science.

  • 6
    Sadly, you can probably only get a "random selection" over those you are able to find and list, not the entire universe. Don't forget to consider sample bias in whatever you do. But if you can get such a list, just index them and use a random number generator.
    – Buffy
    Mar 13, 2022 at 20:07
  • Many (most?) journals have a consistent format for their DOIs, you could try this: crossref.org/documentation/retrieve-metadata/xml-api/… but you'd have to identify the journals and their formats by hand, AFAIK Mar 13, 2022 at 20:14
  • @AzorAhai-him- I have read an answer on this site that DOI formats are not necessarily consistent over time for a given journal, I cannot find it at the moment, though. Mar 13, 2022 at 20:42


You must log in to answer this question.

Browse other questions tagged .