HIVEBEAT — A highly interactive visualization environment for broad-scale exploratory analysis and tracing: VAST 2012 Mini Challenge 1 award: Honorable mention for comprehensive visualization suite

The VAST Challenge 2012 deals with large data network analysis. The challenge scenario is centered around the computer network of the fictitious Bank of Money (BoM). BoM operates 888,977 computers which are geographically distributed on a fictitious landmass. The provided data has a hierarchical structure grouping machines in business units, facilities, machine classes and machine types. For each machine, up to 192 status logs in a two day time period were provided. These logs report on three status attributes: policy, activity and number of connections. The main task of Mini-Challenge 1 was to highlight up to five anomalies in the massive data set. We address this challenge by presenting HIVEBEAT, a highly interactive visualization environment for broad-scale exploratory analysis and tracing. It offers eight interactive views which visualize different aspects of the data, support brushing and linking and are complemented with time-dependent and sequence-based filtering. Some of these components build on ideas from earlier works [2, 3, 4]. With HIVEBEAT, we were able to identify five anomalies and justify them with plausible hypotheses.