Motivation and research in architectural intelligent tutoring

It is well‐known that personalised and adaptive training, such as from a human tutor, is dramatically more effective than traditional classroom training (Bloom, 1984; VanLehn, 2011). Due to a variety of reasons, however, tutoring systems are not yet ubiquitous within the training market. The US Army Research Laboratory is working to address this problem and has recently published a series of research vector outlines, which guide research in the various areas. The research within the architectural vector naturally exists to support the other vectors and to investigate, standardise, componentise, and commodise the processes and functions of the various tutoring system aspects. This paper serves as an expansion and companion to the similarly named 2015 International Defense and Homeland Security Simulation Workshop paper and yet‐to‐be‐published ARL architectural research plan, expanded in order to discuss the progress made to date, clarify the role of the architecture in the research, and discuss some of the advantages of a unified system as part of measuring training effectiveness and overall system improvement.