Application of Adaptive Models to Information Selection in C3 Systems

This report describes research and development centered on the demonstration of an on-line adaptive model for automatically selecting information in a command, control, and communication (C3) system. Rationale for application of the model is built upon a review of psychological literature concerning human performance in specifying information requirements, and in acquiring and utilizing information for military decision making. Based on a multi-attribute decomposition of information messages, the model selects information for an individual user according to his observed information preferences in response to specific situational requirements. An additional algorithm reduces the size of a selected information set by dynamically pruning relatively low-utility items. The model was implemented for a simulated ASW tracking task, and was systematically evaluated in terms of both its intrinsic performance and the performance of an expert operator working with it. The results demonstrated the capability of the model to adapt to varied individualized information seeking strategies, and to subsequently automate the selection of information appropriate to those strategies.