This chapter describes the use of what has been implemented, a
system for data mining.
In particular, we describe the two programs called RGEN and RCLASS\
and the user interface which has been made to simplify and integrate
their use. The integrating part is described in Section 
.
RGEN is a program which creates rules from a data set
using Rough Sets principles. With each rule, it associates a measure of how
good that particular rule is, and how many objects in the training set
it covers. RGEN was created by Jon Petter Hjulstad, and it is
described in [Hju96]. We have made some minor changes, which
are described in Chapter . The use of RGEN is
described in Section
The program RCLASS is a classification program. It uses rules created
by RGEN, and classifies objects according to these. With conflicting
rules, there are several ways of making the classification, and the
user may choose which method to apply. How to use RCLASS is described
in Section , and implementation issues of the program
are described in Chapter .
It is possible to use both RGEN and RCLASS directly without the
graphical user interface. This could be useful if one for
instance only has a text terminal available. It also allows for the
possibility of running test scripts and batch jobs, running the
programs over and over again or at certain times. Command line
arguments are described in subsections and .