Crowdsourced Acquisition of Models of Learning Transfer Strategies (CRAM-LESS)

Navy talent management must evolve to keep pace with the demands of new technology, new battlefields, and new enemy tactics in the face of budget reductions. It must fully utilize the existing skills of Navy personnel, and it must optimize the detailing and subsequent training of personnel to reduce costs and maximize job effectiveness. Transfer is key to talent management; optimal talent management requires understanding which skills can transfer between positions, the training needed to transfer, and which training is unnecessary. To optimize talent management for the Navy, we propose to develop and evaluate a system for Crowdsourced Acquisition of Models of Learning Transfer Strategies (CRAM-LESS). The CRAM-LESS system features: (1) a scientifically grounded model of learning and skill transfer that includes procedural tasks with multiple skill types, such as psycho-motor, perception, and cognitive skills; (2) a method and software infrastructure to crowdsource skill and transfer models; (3) probabilistic reasoning algorithms to create robust models of transfer from noisy crowdsourced data; and (4) integration with advanced and mature optimization algorithms to optimize talent use, maximize career trajectories, and minimize training requirements.