Network Situational Awareness

The system we propose to design and develop, GraphExplorer, will extend previously developed network learning toolkit algorithms to enable a user to interactively explore hypotheses about the structure of large graphs. The system will be an interactive tool, where a user will be able to explore a graph by having the system identify and highlight parts of the graph that are of interest. The user can then choose to collapse the graph using various transformations or guide the system to find other patterns that might be of interest. Generally, machine learning (ML) tools for network learning are not interactive and do not allow the user to explore the data. ML tools (including both supervised and un-supervised systems) are often designed to be run once per data set and return answers in the form of predictions or clusters. The system we propose to build will put the user in the driver's seat and allow for interactive exploration of the data. To this end, we propose to design and apply statistical machine learning technology for detection of groups and patterns within networked data to help better understand the information contained within the graph structure.