METAD: Monitoring Entites for Trend Analysis and Data mining

Our ultimate goal in this project is to develop a methodology and a system, METAD, for analyzing entity information from multiple information sources and assisting analysts in finding suspicious entities in these data.  The system will be able to handle heterogeneous information sources and link entities across these sources into an entity representation that can be mined and analyzed.  The analysis that it will support includes  techniques such as trend analysis, guilt-by-association, as well as social network analysis techniques.   We note that we already have a state-of-the-art entity representation framework in our EntityBase architecture as well as a state-of-the-art relational learning toolkit, NetKit, which includes technology such as guilt-by-association. BENEFIT: By combining new technology for aggregating data with state-of-the-art data mining methods, METAD can vastly simplify this task, and in doing so, open up significant new opportunities for automated monitoring and data mining.  Specifically, the METAD system will provide the framework for acquiring and aggregating real-world dynamic data and the foundation for analyzing such data, facilitating the development of next-generation analysis and mining technologies.  In addition, this effort will provide the foundation for METAD to be developed into a product that can be applied to a variety of domains both in the commercial sector as well as in the intelligence community.