1.2 Knowledge dIsCovery In databases 5
dataset should be used instead of the entire dataset to reduce the time needed
for data mining. The training dataset is obtained during pre-processing, and
it often contains useful information for data analysis. This dataset can also be
used by the data-mining algorithms as a model of the raw data.
Once a certain amount of knowledge is generated by the data-mining
stage, the generated pieces of knowledge have to be converted into a form
more suitable for interpretation by end users in a post-processing stage. The
main purpose of post-processing is to synthesize the generated knowledge
into useful and usable information for strategic decision making by end
users. During the post-processing stage,