,
Supporting multi-dimensional vague queries by using visual filtering
Abstract Proceedings of 34th Annual Conference of the German Classification Society (GfKl 2010), стр. 137. Karlsruhe, (21-23 jul 2010)
Аннотация
Various business processes lean on the knowledge of experts and their
abilities to interpret complex data sets in order to support decision making. We propose a radial user interface which supports unintended, but highly interesting user queries to support system users in a more natural way of interaction. Our interface supports context sensitive parameters by default but enables the users to
personalize these hotspots by adding, deleting, or editing the parameter set. The given parameters for a dataset spans a multi-dimensional query which is visually represented by central focus and which can be affected by the user changing single weight values in order to increase or decrease the importance of some parameters.
These con guration settings have a direct influence on the underlying information. Visualized as data points in the background of the interface they are positioned according their relevance towards a current query con guration. Therewith to achieve a manageable visualization we cluster the data either in the case the number of
data elements exceeds a threshold or the query configuration enforces this. So the granularity level can be system-suggested but is finally controlled by the user.
