Visual Feedback in Querying Large Databases

In this paper, we describe a database query system thatprovides visual relevance feedback in querying large databases.The goal of our system is to support the query specifi-cation process by using each pixel of the display to representone data item of the database. By arranging and coloring thepixels according to their relevance for the query, the usergets a visual impression of the resulting data set. Using slidersfor each condition of the query, the user may change thequery dynamically and receives immediate feedback by thevisual representation of the resulting data set. By using multiplewindows for different parts of a complex query, the usergets visual feedback for each part of the query and, therefore,will easier understand the overall result. The system may beused to query any database that contains tens of thousands tomillions of data items, but it is especially helpful to explorelarge data sets with an unknown distribution of values and tofind the interesting hot spots in huge amounts of data. The directfeedback allows to visually display the influence of incrementalquery refinements and, therefore, allows a better,easier and faster query specification.

Authors: Keim D. A., Kriegel H.-P., Seidl T.
Published in: Proc. IEEE Int. Conf. on Visualization 1993, San Jose, California
Publisher: IEEE Computer Society - Washington, USA
Sprache: EN
Jahr: 1993
Seiten: 158-165
Konferenz: Visualization
Typ: Tagungsbeiträge
Forschungsgebiet: Exploration of Multimedia Databases