Virtual Intelligent Collaboration for Task Organization and Resource Scheduling

The need to make rapid and well-informed decisions about custom task organizations or ad-hoc problem-focused staffing is shared by organizations large and small, but is especially critical to the military. The problem has components of (1) resource discovery and characterization, (2) task analysis and needs identification, and (3) flexible interactive and justifiable dynamic scheduling. We propose to develop a system to support Virtual Intelligent Collaboration for Task Organization and Resource Scheduling (VICTORS). VICTORS will combine (a) distributed text mining and data query, (b) social network exploitation, (c) task model acquisition, (d) policy representation and reasoning, and (e) flexible interactive scheduling tools. During Phase I, we propose to address the first three of these technologies, in descending order of emphasis. Following on requirements analysis and technical surveys, we will develop a limited proof-of-concept prototype that illustrates some of the most promising techniques, focusing on lowering the costs and raising the precision of information extraction techniques. Our Phase I results will include a preliminary Phase II system design and work plan, pointing the way towards a complete solution to the team formation problem.