![]() |
Volume 46, Number 5
(September - October 2001) |
Key words: computational intelligence, mass spectroscopy, rule-base reasoning
In this paper (part of sequence devoted to automatic identification
of organic substructures [1,2]) a methodology of searching for optional
classifiers for selected aromatic fragments embedded in organic molecules
is briefly described. The developed methodology uses low-resolution mass
spectra and employs computer program SCANKEE to create the databases of
mass spectra and to search them to create spectrum-substructure correlation
tables, and finnaly to convert automatically these tables into the rules
database which enable effective concluding.
back
to table of contents