konstanz.de

konstanz.de Real-time analysis of data streams has become an impor-tant factor for success in many domains such as server and system administration, news analysis and finance to name just a few. Introducing real-time visual analytics into such application areas promises a lot of benefits since the rate of new incoming information often exceeds human perceptual limits when displayed linearly in raw formats such as tex-tual lines and automatic aggregation often hides important details. This paper presents a system to tackle some of the visualization challenges when analyzing such dynamic event data streams. In particular, we introduce the Event Visual-izer, which is a loosely coupled modular system for collect-ing, processing, analyzing and visualizing dynamic real-time event data streams. Due to the variety of different analy-sis tasks the system provides an extensible framework with several interactive linked visualizations to focus on different aspects of the event data stream. Data streams with logging data from a computer network are used as a case study to demonstrate the advantages of visual exploration