Walks in cyberspace: Improving web browsing and network activity analysis with 3D live graph rendering

Web navigation generates traces that are useful for Web cartography, user behavior analysis (UEBA) and resource allocation planning. However, this data still needs to be interpreted, sometimes enriched and appropriately visualized to reach its full potential. In

this paper, we propose to explore the strengths and weaknesses of standard data collection methods such as mining Web browser history and network traffic dumps. We developed the DynaGraph framework that combines classical traces dumping tools with a Web application for live 3D rendering of graph data. We show that mining navigation history provides useful insights but fails to provide real-time analytics and is not easy to deploy. Conversely, mining network traffic dumps appears easy to set up but rapidly fails

once the data traffic is encrypted. We show that 3D rendering allows to highlight navigation patterns for a given data sampling rate.