- Exploratory data mining is fundamental to fostering
an appreciation of complex datasets. For large
and continuously growing datasets, such as obtained
by regular sampling of an organisation's communications,
the exploratory phase may never finish.
This paper describes a methodology for exploratory
data mining within an organisational communications
dataset. A model of support for knowledge discovery
is described in conjunction with a communications
based concept hierarchy. This is then used as
the basis for a set of visualisations. The intention
of supporting visualisations in this way is to establish
a sound set of requirements for the representation
of communications data. The visualisations provide
several interconnected representations of the data, as
well as support query and drill-down into a dataset.
It is suggested that this interaction with the dataset
facilitates an appreciation of the data which precedes
and shapes knowledge discovery. A communications
analysis example is developed using the visualisations
within the context of exploratory data mining.
(from crpit.com)
(local if available)
