Jeremiah T. Folsom-Kovarik

Tractable POMDP representations for intelligent tutoring systems

With Partially Observable Markov Decision Processes (POMDPs), Intelligent Tutoring Systems (ITSs) can model individual learners from limited evidence and plan ahead despite uncertainty. However, POMDPs need appropriate representations to become tractable in ITSs that model many learner features, such as mastery of individual skills or the presence of specific misconceptions. This article describes two POMDP representations—state queues and observation chains—that take advantage of ITS task properties and let POMDPs scale to represent over 100 independent learner features. A real-world military training problem is given as one example. A human study (n = 14) provides initial validation for the model construction. Finally, evaluating the experimental representations with simulated students helps predict their impact on ITS performance. The compressed representations can model a wide range of simulated problems with instructional efficacy equal to lossless representations. With improved tractability, POMDP ITSs can accommodate more numerous or more detailed learner states and inputs.

Adaptive Perceptual Training in a Virtual Environment

The United States military’s strategic position is evolving, and as a result, the Services are emphasizing the importance of sociocultural pattern recognition, sensemaking in ambiguous urban contexts, and understanding of patterns of life. In fact, military personnel at increasingly lower echelons are expected to possess these nuanced psychosocial perception and decision-making skills. To facilitate training of these complex competencies, the authors are developing a Virtual Observation Platform, an immersive virtual environment designed to adaptively train US Marine Corps personnel in sustained observation, sociocultural pattern recognition, anomaly detection, and other perceptual–cognitive skills. This paper briefly describes the purpose of the system and then covers its adaptive instructional tailoring in detail. The Platform’s adaptive features include information quality/quantity manipulation and instructional scaffolding in the form of communications from a virtual squad (i.e., peers) that are intrinsic to the narrative of the scenario.

ITS + SBT: A review of operational situated tutors.

Situated tutors combine features of intelligent tutoring systems with simulated environments. A construct definition for situated tutors was newly developed, and, in this article, we use it to classify an extensive review of operational instructional technologies in order to identify a robust set of situated tutors. We document 86 situated tutors, half of which directly support military training. Initial empirical evidence from these systems suggests that situated tutors may outperform traditional simulations or intelligent tutors. As additional evidence is gathered, we hypothesize that situated tutors will also demonstrate efficiency and effectiveness in cognitive readiness instruction.

Instrumenting a Perceptual Training Environment to Support Dynamic Tailoring

Simulation-based practice environments would be more valuable for learning if they supported adaptive, targeted responses to students as they proceed thru the experiences afforded by the environment. However, many adaptation strategies require a richer interpretation of the student’s actions and attitudes than is available thru the typical simulation interface. Further, creating extended interfaces for a single application solely to support adaptation is often cost-prohibitive. In response, we are developing “learner instrumentation middleware” that seeks to provide a generalized representation of learner state via reusable algorithms, design patterns, and software.

AI Challenge Problem: Scalable Models for Patterns of Life

This article focuses on the problem of patterns of life (POL), which emerge from human social systems. It describes conflicting requirements for this problem and potential AI solutions.

Using Stories to Deepen Shared Human-Computer Understanding of PMESII Systems

Humans tend to represent and to understand the world in terms of stories, while computer reasoning tends to require formal, mathematical representations. This paper describes a research prototype that enables computers to parse human stories and use collections of those stories to inform causal modeling of political, military, economic, social, infrastructure, and information (PMESII) systems. We introduce the need for causal modeling, the approach we have taken in implementing an initial proof-of-concept and the results from pilot testing of the software that illustrates functional capabilities and opportunities for deepening story-based computer interpretation of stories.

Humans as the Strong Link in Securing the Total Learning Architecture

This paper describes a proposed approach, centered on human factors, for securing the Total Learning Architecture (TLA). The TLA, which is being developed for the United States Department of Defense, will rely on large stores of personal data that could be targeted by sophisticated adversaries. We describe the TLA and its envisioned users at a fairly high level before describing expected classes of attacks against it. We then examine existing and proposed controls that, if properly managed, should allow users and service providers to significantly reduce the risks to the system.

Evaluating Instructor Configurability for Adaptive Training

Adaptive training technologies offer the promise of more individualized and effective training. However, these technologies increase the complexity of training systems. They also potentially can limit the ability of instructors and curriculum developers without technical skills to customize and to adapt training to specific instructional demands. This paper describes a methodology for verifying that an adaptive training system is configurable and responsive to specifications encoded by instructors. That is, when an instructor desires for the training system to respond in particular ways in particular circumstances, how readily can the adaptive training technology be used to execute that specification? The paper describes a verification methodology and its application to a dynamic adaptation capability in a desktop-based simulation-training prototype.