The objective of this grant is to address research issues related to semantic interoperability, data mining and security and authorization in the multi-agency domain of border control. Partners are the Immigration and Naturalization Agency, the Customs Service, and the Coast Guard. In addition to the research issues that will be addressed in this proposed work, the development of a proof of concept in the form of a prototype will be undertaken. Specifically, in the area of semantic interoperability, the team focuses on providing expressive common language among information sources, automatic mapping, and the ability to dynamically adapt to changing environments. Furthermore, automatic text analysis and text mining techniques will be developed to automatically, rather than manually, build the knowledge base. The team addresses some of the challenges encountered when sharing information in a secure manner, including reconciling the heterogeneity among multiple security policies, resolving semantic heterogeneity among security policies, translation of coalition level policies to implementation level and vice versa, and support of delegation in coalition based systems. Also addressed will be issues related to data mining to build an alert system. This includes developing classification/segmentation models and merging of different local models to a global model for outlier analysis, real time data mining, and providing iterative and interactive visualization. The multidisciplinary research team, which comes from multi-institutions is augmented by a set of domain experts in several related areas, including homeland security, border control, global logistics, transportation security, foreign terrorist tracking, as well as domestic and multinational investigations. These domain experts provide guidance on the functional requirements of the system and feedback on the prototype system as it progresses through its development.
