Monitoring and analysis of streaming data, such as social media, sensors, and news feeds, has become increasingly important for business and government. The volume and velocity of incoming data are key challenges. To effectively support monitoring and analysis, statistical and visual analytics techniques need to be seamlessly integrated; analytic techniques for a variety of data types (e.g., text, numerical) and scope (e.g., incremental, rolling-window, global) must be properly accommodated; interaction, collaboration, and coordination among several visualizations must be supported in an efficient manner; and the system should support the use of different analytics techniques in a pluggable manner. Especially in web-based environments, these requirements pose restrictions on the basic visual analytics architecture for streaming data. In this paper we report on our experience of building a reference web architecture for real-time visual analytics of streaming data, identify and discuss architectural patterns that address these challenges, and report on applying the reference architecture for real-time Twitter monitoring and analysis.
Visual display of information in data mining can support successful knowledge discovery. An experiment was conducted to identify parameters that affect the detection of cause-and-effect relations in time series data in a visual data mining environment. Accuracy of performance and the frequency of tool usage were measured as a function of visual properties of the cause-function and information processing styles. Performance accuracy differed between participants with different cognitive styles. Participants with high analytic cognitive style were better able to detect cause-and-effect relations through the investigation of visual and more global properties of the displayed data. Visual properties of the data affected users with high analytic and low experiential cognitive styles similarly and had no direct effect on accuracy. Participants with different levels of cognitive style differed in tool usage, indicating diverse approaches to solving the experimental task. The results point to the need to consider the effects of user characteristics and properties of the displayed data when designing visual data mining environments that are based on intense interaction of users with complex graphical displays.