An advanced network visualization system for financial crime detection

Abstract

We present a new system, VISFAN, for the visual analysis of financial activity networks. It supports the analyst with effective tools to discover financial crimes, like money laundering and frauds. If compared with other existing systems and methodologies for the analysis of criminal networks, VISFAN presents the following main novelties: (i) It combines bottom-up and top-down interaction paradigms for the visual exploration of complex networks; (ii) It makes it possible to mix automatic and manual clustering; (iii) It allows the analyst to interactively customize the dimensions of each cluster region and to apply different geometric constraints on the layout. VISFAN also implements several tools for social network analysis other than clustering. For example, it computes several indices to measure the centrality of each actor in the network.