Re: No subject

peter Murray-rust (
Sat, 10 Feb 1996 23:02:36 +0000 (GMT)

On Sat, 10 Feb 1996, Lesley West wrote:
> OK, here's where we need to clarify if we need to assign keyword
> structure. MESH and other nomenclatures have precise structures which
> may not fit into virtual hyperglossaries unless we think about them at
> this stage. For example, do we need to structure the preferred term
> compared to synonyms or abbreviations? If you take 'acid', then link the
> 'sub-directory' amino acid, how does the reader (not curator) know
> this is a Higher Term with many lower terms below it (in MESH the
> abbreviation HT is used etc)?

Firstly, gentle reader, don't panic. If you don't understand this, don't
worry. Lesley and I can thrash it out in a pub on Wednesday :-)

I see the PPS96 glossary as a collection of, say, 300 terms. Some of
those will have higher terms (e.g. tyrosine ISA aminoacid). All higher
terms should therefore have lower terms. Therefore tyrosine should
somewhere have a field <term type="superordinate concept">aminoacid</>.
Similarly 'aminoacid' should have a set offields like:
<term type="subordinate concept">tyrosine</>
<term type="subordinate concept">tryptophan</>

IMO it is the job of the software to reveal any inconsistencies here and to
generate a self-consistent structure. At the present, don't worry and we
will provide either a simple template or a tool to load these into the
glossary (or both).


Peter Murray-Rust, Glaxo Research & Dev. (; (BioMOO: PeterMR)
Birkbeck College,, CBMT/Daresbury,