This paper describes an advanced decision support
system for scheduling that is based on both, extensive
knowledge about scheduling algorithms, and a new software engineering paradigm developed by the
FAW. This paradigm uses an elaborate compromise
between object-orientation, constraint processing, and
functional capabilities, and supports distributed computing.
The system has the capability to generate a model
of itself. This Eigenmodel is a directed graph G representing
data structures (nodes of G) and algorithms
(edges of G). The methods developed for scheduling
problems can, on a “higher” level of abstraction, also
be applied to the Eigenmodel itself. This can then be
used for performance analysis and learning aspects
which will eventually provide the user with valuable
information about the functioning and performance of
the system, thus contributing to the envisioned new
generation of decision support systems.