The UK spent several million pounds on refining the Read Codes. They got hundreds of experts to sit around and gaze at their navels and produce terms, and argue about how to place them in a hierarchy. It kinda seemed like a good idea at the time, it really did.
Knowing what I know now tho, what I really want is to get my hand on the results of this exercise, which seems to be available for free here. To me, this is practical knowledge engineering.
Less simple, but very doable, is to take the flat list of terms in this subset and look at them in the context of the full UMLS/SNOMED semantic network. Using the hierarchy and other relationships it should be possible to rapidly define usable filters and grouping systems to start bringing some order to the list.
Knowing what I know now tho, what I really want is to get my hand on the results of this exercise, which seems to be available for free here. To me, this is practical knowledge engineering.
Less simple, but very doable, is to take the flat list of terms in this subset and look at them in the context of the full UMLS/SNOMED semantic network. Using the hierarchy and other relationships it should be possible to rapidly define usable filters and grouping systems to start bringing some order to the list.
No comments:
Post a Comment