I wanted to get more information about the way PHOEBE makes recommendations on concepts, e.g. is it based on similar wording of concepts that exist in the 22 databases that are integrated? I was not able to find a documentation or a published paper regarding this. Would @aostropolets be the right person to provide some further information on it?
In addition to that, I was wondering whether it is possible to run PHOEBE on concepts that are not standard codes. So far I was only successful at using it when I used standard SNOMED concepts for conditions.