An Adaptive and Distributed Approach to Developing and Maintaining an Integrated Search and Discovery Portal

We propose an innovative approach to supporting the Army's Homeland Security (HLS) mission that turns the diversity of perspectives that exist amongst largely autonomous HLS communities in DoD and non-DoD organizations into an exploitable strength. In particular, we propose to develop an approach to the development and maintenance of a search and discovery portal that exploits an eclectic set of techniques in order to achieve functional and semantic interoperability amongst these communities without forcing participants to conform to the type of oppressive monolithic semantic schema that has led to the demise of so many corporate knowledge management efforts over the past decade. Rather, the proposed system, Vista, will employ an adaptive distributed paradigm where individuals and groups can manage their local information in a manner that fits their respective workflow patterns and Vista will exploit this diversity in a machine learning based bootstrapping approach to meaning negotiation. The resulting system will be capable of delivering personalized and context-aware information access and awareness in support of the Army's information superiority needs. Phase I research and the development and evaluation of a limited prototype will provide a solid foundation for the complete implementation of Vista in Phase II and its commercialization.