Creating search strings with AI

AI has clear limitations as a working tool in academic practices. This is partly due to the fact that it is difficult to have an overview of how AI operates. To support searches, AI provides random hits, it invents articles and references, while the results cannot be reproduced. Using AI as a support tool in specific sub-tasks like creating search strings can be far more useful.  In this workshop, we will look closer at how AI can be used to create or improve search strings that can be used in databases relevant to your field.  This can be used in literature review, kappa etc. 

Course Content  

The workshop will first introduce how a search is set up, without AI. The introduction will demonstrate ways to refine and narrow searches in search strings using Boolean operators (AND, OR, ++), wildcards and search syntax. Then we will show how you can use natural language to get Copilot to suggest search strings where these techniques are used.   We take a closer look at what you should be aware of, what instructions (prompts) work and how you can check that the search strings are well designed.  

Design 

The course takes the form of a workshop, and plenty of time will be set aside to build your own search strings and test these in preferred databases. Feel free to indicate which databases for research literature are most relevant to you when you sign up, and we will try to use these as examples.    

Duration: 

2 hours including intermission. The room is reserved for an optional extra hour.