In sparse temporal datasets it is important to have a querying system to detect patterns. This general task raises questions about how to model the dataset, how to query the dataset and how to visualize the results.

In high intensity medical environments such as an Emergency Department (ED), clinicians should become productive immediately upon arrival. The less intuitive a system is, the more time it takes for training and the more prone it is to human factors breakdown. In such environments, an intuitive visual interface for their temporal dataset is imperative.

This paper presents an interactive interface that builds on previous research in visual and dynamic query systems. The Patternfinder system has been applied to a temporal data set in the medical domain from Washington National Hospital Center.

The temporal dataset was extracted from a relational database with time stamps for each tuple. Using PatternfinderĄ¯s visual query interface, doctors or clinicians can create most commonly used queries. The visualization panel will display the result set in a timeline that gives overview and detail of the retrieved result.