Hi Kean, I think that it excludes terms that start with "no" because the usual desired behavior is to rely upon one of the negation engines to determine that status.
That being said, missing things like "non Hodgkin's" is definitely not a desired behavior. I will look at the code and see if I can determine what happened there. Thanks for reporting, Sean ________________________________ From: Kean Kaufmann <k...@recordsone.com.INVALID> Sent: Tuesday, April 16, 2024 10:42 AM To: dev@ctakes.apache.org <dev@ctakes.apache.org> Subject: Custom dictionary no-"no" [was: Re: PREFTERMs not included in UMLS rare-word dictionary?] [EXTERNAL] * External Email - Caution * Hi Sean, I ran the dictionary creator tool from ctakes 5.0.0, and am happy to see that preferred texts are now also lookup texts -- thank you! However, I also note that almost all terms starting with the letters "no" are now omitted, except for a few starting with "no fh : ". From my custom dictionary, that's about 21K terms missing, including common ones like "no appetite", "nocturia", "nodule", "non hodgkin lymphoma", "nondisplaced intertrochanteric fracture", "normal pressure glaucoma", ... Is this filtering expected? Is there a way for the user to control it? Thanks as always, Kean On Wed, Dec 6, 2023 at 5:22 PM Finan, Sean <sean.fi...@childrens.harvard.edu.invalid> wrote: > Hi Kean, > > I can't think of a good reason for preferred text to not also be a lookup > text. It sounds like you might have uncovered a flaw in the dictionary > creator tool. > > Time for a rebuild with the 5.0 release ... > > Thanks for the report, > > Sean > > ________________________________ > From: Kean Kaufmann <k...@recordsone.com.INVALID> > Sent: Wednesday, December 6, 2023 4:12 PM > To: dev@ctakes.apache.org <dev@ctakes.apache.org> > Subject: PREFTERMs not included in UMLS rare-word dictionary? [EXTERNAL] > > * External Email - Caution * > > > Hi Sean and fellow Fast Dictionary Lookup fans, > > I notice that the UmlsJdbcRareWordDictionary doesn't seem to index terms > from PREFTERM, only CUI_TERMS. > ...
