Interactive refinement of filtering queries on streaming intelligence data

Abstract

Intelligence analysis involves routinely monitoring and correlating large amount of data streaming from multiple sources. In order to detect important patterns, the analyst normally needs to look at data gathered over a certain time window. Given the size of data and rate at which it arrives, it is usually impossible to manually process every record or case. Instead, automated filtering (classification) mechanisms are employed to identify information relevant to the analyst's task. In this paper, we present a novel system framework called FREESIA (Filter REfinement Engine for Streaming InformAtion) to effectively generate, utilize and update filtering queries on streaming data. © Springer-Verlag Berlin Heidelberg 2006.