Establishing Conclusive Proof in Forensic Data Analysis

Gabriel Hopmans Morpheus, Netherlands

Description

SBV Forensics is a specialist in fraud in cases similar to the famous Enron scandal. Its core business is investigating incidences, irregularities, illegalities and/or criminal conduct in the field of financial administration. Some of this work involves the analysis of huge amounts of data.

In such cases, analysts, forensic accountants and related business users spend much of their time looking for evidence and uncovering incriminating relationships in information systems and environments. In many traditional forensic applications it is possible to visualize networks of relations and perform complex searches, but the real domain knowledge cannot be represented.

To counter this, SBV Forensics uses Topic Maps to connect and leverage existing indexes and structured data sources in order to establish more conclusive proof, and to combine this pre-existing data with the knowledge of domain experts: lawyers, liquidators and (forensic) accountants. These people know a lot about relevant indicators of behavior and they make a lot of assertions and assertions about assertions, all of which can be captured in a topic map. In particular, the necessary contextualization (for example, who made which assertion) is easily represented in the topic map.

In one particular case the main challenge was to find evidence of fraud in millions of e-mails. To handle this problem the core concepts mentioned in the formal injunction formed part of a topic map which was used by the fraud analyst in conjunction with a Topic Maps-based search tool to support his research. As a result, new facts were uncovered that could be used to strengthen the allegations of the plaintiff.

This presentation discusses some of these cases using anonymized data and shows the benefits that Topic Maps has provided.

Slides

http://www.topicmaps.com/tm2008/hopmans-kruijsen.pdf

http://www.topicmaps.com/tm2008/hopmans-kruijsen.ppt

Other presentations:

TopicView and Terrorist Ontologies (Topic Maps 2007)