Innovative Filtering Techniques and Customized Analytics Tools

Harald Bosch, Julian Heinrichm, Benjamin Höferlin, Markus Höferlin, Steffen Koch, Christoph Müller, Guido Reina, Michael Wörner

Description

The VAST 2009 Challenge consisted of three heterogeneous synthetic data sets organized into separate mini-challenges with minimal correspondence information. The challenge task was the identification of a suspected data theft from cyber and real-world traces. The grand challenge required integrating the findings from the mini challenges into a plausible, consistent scenario. A mixture of linked, customized tools based on queryable models and rapid prototyping as well as generic analysis tools (developed in-house) helped us correctly solve all of the mini challenges.

A collaborative analytic process was employed to reconstruct the scenario and to propose the correct steps for the reliable identification of the criminal organization based on activity traces of its members. piciously. The social network analysis provides information about who in the criminal network knows at least two other persons located in their vicinity. We inferred plausible mappings from these persons to their role in the organization and formulated ways to confirm these hypotheses, e.g. by checking the embassy roster. The following sections describe our MC tool sets to discover and relate
findings.

Video