Some journals require keywords besides the abstract to 'describe' the published article, some don't. See for instance with keywords http://www.sciencedirect.com/science/article/pii/S0049089X16301788 and without keywords http://onlinelibrary.wiley.com/doi/10.1111/ajps.12313/full
I was wondering whether there are guidelines or recommendations how to choose such keywords? For instance, has each journal a list of keywords that you can use or recommend scientific societies/associations a certain standard of 'how to set up meaningful keywords'? Or do paper authors just make up their own keywords?

4 Answers 4


Actually, very few journals have recommendations on keywords, it is usually left to the authors to decide on the keywords. Most of the times, it is easy to come up with the keywords for papers with narrow scopes, but broader ones are troublesome.

Here is a recommendation that you can consider.

Telescope and Microscope Method:

Let us say you have decided on an arbitrary number of x keywords - divide them into five equal parts as follows:

  • 20% of them should cover the wide scope of your paper (pharmacoeconomics, therapeutics, cost-benefit ratio)
  • 20% should narrow down to the most frequently used words in your article
  • 20% should be words that DESCRIBE your study and its conclusions
  • 20% should be rare words, acronyms and obscure references, unique statistical or methodological methods, eponyms.
  • 20% should be SEO oriented: That is, searchable keywords you anticipate your potential readers to use to locate your paper.

Hope this helps.

  • Yes, I actually found in my discipline (Political Science) no recommendations at all and also no reason why some journals use keywords while others don't. So, your suggested divide in parts is quite interesting. I will definitely consider this!
    – Stefan_W
    Commented Jun 24, 2017 at 13:15

The following are not exactly a set of keywords, but are somewhat related:


Since keywords are meant to be used as search indices for certain topics, allowing authors to generate their own keywords makes little sense. For all journals I know, keywords have to be selected from a list published by the journal. Some journals have adopted common keyword lists (for example most journals in astrophysics use this list).

Keywords in academia play almost exactly the same role as tags on SE sites (except that author cannot invent new keywords). In particular, their selection by the authors follows the same principle: find the most suitable set of keywords for the given study, such that a keyword search brings up the most relevant articles.

Keywords are perhaps not as important now as they used to be since electronic word searches on title, abstract or the whole article.

  • Thanks for the link to the list and it sounds reasonable. I was wondering hom important these keywords still are (in times of google scholar etc.). I guess, google scholar won't give a list of relevant articles based on keywords, but probably on broader algorithms for the whole article or at least for the abstract of it, right?!
    – Stefan_W
    Commented Jun 24, 2017 at 13:17

It might be worth looking into the (US) National Library of Medicine's "Medical Subjects Heading" program, starting here.

These keywords, which they call "terms", are assigned to PubMed-indexed articles by the NLM, rather than the authors, but nothing is stopping authors and journals from using them too; in fact, they suggest that!

This would be miles better than the keyword system used by most biomedical journals, which is either a) a half-baked collection of terms, focused on a long-departed editor's interests or b) a text box, a wing, and a prayer.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .