For my current project https://www.researchgate.net/project/The-use-of-international-cooperation-in-faculty-development-of-health-professions-education-a-rapid-scoping-review I have been writing a study protocol Research Proposal STUDY PROTOCOL | The use of international cooperation in fac...
At the end of the process I was looking for some fitting KEYWORDS as to improve meta-indexing of the protocol.Following the guidance by JBI: "Ideally these words should be different to those that appear in the title and abstract. These are for the purposes of meta-data and indexing, and not related to the search strategy." plus reading some guides on how to choose Keywords, I came up with the MESH-term synonyms of my own main terms. To that I then added some "concept" keywords from studies I am aiming for in my proposed rapid scoping review.
The problem I am having is, that I am aiming for a multitude of studies more, which are not solely represented by that array of keywords as e.g. studies with a developmental aspect or a service learning focus or evolving around internationalisation of curricula or something like a sabbatical for personal development! SO my question is, is there any theoretical tool for text/concept generation of keywords OR maybe are there even already some kind of software tools to that regard? Any directing will be highly appreciated! And for the ones, wanting to have a short look, I would be very grateful for any hints and tips on the actual protocol.
The Keywords I came up with in the protocol are:
health occupations teaching; higher education; academic internationalisation; staff advancement; transnational collaboration
thanks in advance
....the JBI quote in the question is from the SUMARI protocol template available at https://jbi.global/sites/default/files/2020-07/SUMARI_Protocol_Template_Scoping_Reviews.docx
If I'm understanding your question right, I've had some success using noun phrase extraction using Python TextBlob (https://www.pythonprogramming.in/how-to-extract-noun-phrases-using-textblob.html).
Python NLTK is another option but involves a lot more reading and trial and error. TextBlob simplifies the process and from my experience gave better results.
Either way, you'll end up with a list of terms that you'll have to manually verify, but the code should do a lot of the work.
... for anybody with the same question, a very useful article on the issue using text-mining to find keywords:
Text mining for search term development in systematic reviewing: A discussion of some methods and challenges
Article Text mining for search term development in systematic review...
hope it helps
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs