Open bmschmidt opened 1 year ago
That sounds cool. My impression when having a look at the MeSH headers (a long time ago) was that they were very specific, but maybe I didn't get the whole picture of the hierarchy. Maybe we can discuss about it sometime in the future and if they are useful, create some coloring based on MeSH for when we update the whole dataset and everything else.
Yeah they're specific, but organized into a hierarchy where COVID-19 is, say, unambiguously in 'diseases' (C) then in either infections or respiratory tract diseases... etc.
I've got pretty good MeSH data parsed out for every pubmed article right now. Not going to try to drop shoehorn this in before launch, but I think there's a good way to use the intrinsic MeSH hierarchy to color articles at multiple levels. Filing as an issue mainly to note for @ritagonmar that I've got some pretty solid parsing code for this stuff right now on the 2023 dataset, and the beginnings of an understanding of the MeSH hierarchy.