Mind the time: Unleashing temporal aspects in pattern discovery

• We propose a new algorithm for Temporal Data Mining that combines more information in fewer patterns.
• The algorithm allows to freely define relations between events.
• A theoretical assessment of runtime and memory consumption is tested with a benchmark.
• We argue how the algorithm can be interactively parametrized and the results be explored by means of Visual Analytics.
• The algorithm and the current interface are tested in a usage scenario.

Temporal Data Mining is a core concept of Knowledge Discovery in Databases handling time-oriented data. State-of-the-art methods are capable of preserving the temporal order of events as well as the temporal intervals in between. The temporal characteristics of the events themselves, however, can likely lead to numerous uninteresting patterns found by current approaches. We present a new definition of the temporal characteristics of events and enhance related work for pattern finding by utilizing temporal relations, like meets, starts, or during, instead of just intervals between events. These prerequisites result in MEMuRY, a new procedure for Temporal Data Mining that preserves and mines additional time-oriented information. Our procedure is supported by SAPPERLOT, an interactive visual interface for exploring the patterns. Furthermore, we illustrate the efficiency of our procedure presenting a benchmark of the procedure's run-time behavior. A usage scenario shows how the procedure can provide new insights.

Our procedure constructs events of irregular length from data values based on conditions. These events are combined to patterns. In this example, a Thursday with unusually low traffic creates patterns that break up the otherwise regular week patterns.Figure optionsDownload high-quality image (250 K)Download as PowerPoint slide