Standard quantitative analyses do not align with the qualitative information. Orthodox searches of the text content of electronic documents do not map the interconnections between actors or the pathways of financial flows. Integrating structured data, such as financial transaction records, with unstructured data, such as emails and attachments, on a common platform makes it possible to use sophisticated analytics that increase the productivity and the power of investigative strategies. This is especially important in financial cases, in which qualitative information (documents) and quantitative information (data) are critical. Trade-based money laundering cases are an example of such cases.
Our methods recognize that when properly analyzed, documents are just another form of data. We use analytics to direct manual investigation efforts on certain activities that have the highest expected value for the case. For example, we use tailored text mining and topic-modeling techniques to identify threads of discussions across documents that help focus investigators on productive areas for review and follow-up. Combining these tools with traditional investigative techniques may significantly reduce the number of false positives and resources needed for the investigation.
Advanced network analytics allow us to map emails or financial transactions between individuals or entities and trace chains of sequential communication and financial flows not easily visible through examination of individual emails or transactions. We create graphics that quickly identify important channels and intermediaries and identify new targets of investigation.
Our experience has shown that new data science tools that combine qualitative and quantitative analytics have important applications to financial investigations and can be a critical element in successfully navigating complex cases.