Latent Semantic Analysis for Assessment and Training of Knowledge and Competencies in Air Force Intelligence Organizations

A highly automated approach is proposed to measure learning and performance of cognitive tasks in Air Force intelligence organizations. We propose to identify key knowledge requirements, Mission Essential Competencies, and training needs in arepresentative intelligence domain; develop a Latent Semantic Analysis (LSA) model of domain knowledge with which we will assess expertise and diagnose specific knowledge deficits among individuals and teams; devise training strategies that take advantageof automated assessment capabilities; and illustrate them ? along with assessment and knowledge acquisition techniques ? in a web-based conceptual prototype. Work in this and subsequent SBIR Phases should produce: reliable, valid, and non-invasivemeasurement and diagnosis of knowledge and competencies in the CAOC or DCGS; improved selection, training, and real-time decision support; improved planning of new intelligence missions supported by a topically organized knowledge base of prior missions.