I'll post a more detailed answer in the next day or two, but at least briefly, when you use RELDEF you are specifying the CUI from which a definition is taken.
So, in the case of RELDEF :: CUI you are simply saying give me the definition of the CUIs I have specified in similarity.pl. This is the natural case. You can also say things like RELDEF :: CUI,PAR which means give me the definitions of the CUIs I have specified, plus those associated with the parents of those CUIs. this is how we created an extended gloss overlaps (where the definition of a CUI includes the definition of the CUI itself plus the definitions of its parents). RELDEF :: TERM means that rather than using a definition, you use the associated terms of the given CUIs in place of a definition. So, that's a start of a response, I'll continue with more in the next day or two. Cordially, Ted On Mon, Jul 28, 2014 at 6:02 PM, [email protected] [umls-similarity] <[email protected]> wrote: > > > Hi > > > The Lesk algorithm computes similarity between glosses of two concepts. As > per the article (Banerjee and Pedersen 2002) an overlap is computed as the > longest sequence of one or more consecutive words. > > > I ran lesk (UMLS3013AA) on a set of 58K pairs using > > > config1 > > SABDEF:: SNOMEDCT > > RELDEF:: CUI > > and almost all pairs had a score of -1. > > > But with > > config2 > > SABDEF:: SNOMEDCT > > RELDEF:: TERM > > has a number of positive scores. > > > I would like to know > > 1. What exactly is the gloss when RELDEF has TERM ? Is it the str column > in the mrconso table of that cui ? > > > 2. As per my understanding of the documentation, CUI in RELDEF implies > that the cui definition is fetched from the mrdef table, which is treated > as a gloss. Is this correct ? If CUI is not specified, what is the gloss ? > In either case i am unable to understand why config1 gets no output. > > The concepts i am working on have been extracted from a dataset associated > with the same disease. I ran similarity measures on them many yield > positive output for similarity. > > > 3. The published paper is for WordNet and the configurations are specific > to UMLS. It would be great if you can provide brief commentary on what > constitutes a gloss for different options in UMLS. > > > Thanks, > > Chaitanya. > > > > > > -- Ted Pedersen http://www.d.umn.edu/~tpederse
