Assistive interfaces, such as recommendation engines, adaptive systems, and intelligent assistants, span diverse methods and disciplines but lack a shared conceptual foundation. This paper models assistance as sequential decision-making under uncertainty between two agents: the user and the assistant. The formalism allows casting assistance as an optimization problem and offers a rich but principled vocabulary to understand the dynamics of assistance. Drawing on Partially Observable Stochastic Games (POSGs) and related models, we: (1) motivate multi-agent over single-agent formulations; (2) adapt POSGs to HCI and clarify their tractability through reductions; (3) propose a two-agent sequential model that unambiguously defines concepts such as adaptation, augmentation, and delegation; (4) illustrate applicability through domain problems and examples; and (5) offer a supporting implementation via a library. These results warrant more attention on decision-theory as a principled yet actionable approach to assistive interfaces.

Abstract Users often browse the web in an exploratory way, inspecting what they find interesting without a specific goal. However, the temporal dynamics of visual attention during such sessions, emerging when users gaze from one item to another, are not well understood. In this paper, we examine how people distribute visual attention among content items when browsing news. Distribution of visual attention is studied in a controlled experiment, wherein eye-tracking data and web logs are collected for 18 participants exploring newsfeeds in a single- and multi-column layout. Behavior is modeled using Weibull analysis of item (article) visit times, which describes these visits via quantities like durations and frequencies of switching focused item. Bayesian inference is used to quantify uncertainty. The results suggest that visual attention in browsing is fragmented, and affected by the number, properties and composition of the items visible on the viewport. We connect these findings to previous work explaining information-seeking behavior through cost-benefit judgments.

Theory-based, or “white-box,” models come with a major benefit that makes them appealing for deployment in user modeling: their parameters are interpretable. However, most theory-based models have been developed in controlled settings, in which researchers determine the experimental design. In contrast, real-world application of these models demands setups that are beyond developer control. In non-experimental, naturalistic settings, the tasks with which users are presented may be very limited, and it is not clear that model parameters can be reliably inferred. This paper describes a technique for assessing whether a naturalistic dataset is suitable for use with a theory-based model. The proposed parameter recovery technique can warn against possible over-confidence in inferred model parameters. This technique also can be used to study conditions under which parameter inference is feasible. The method is demonstrated for two models of decision-making under risk with naturalistic data from a turn-based game.

Decision-theoretic models explain human behavior in choice problems involving uncertainty, in terms of individual tendencies such as risk aversion. However, many classical models of risk require knowing the distribution of possible outcomes (rewards) for all options, limiting their applicability outside of controlled experiments. We study the task of learning such models in contexts where the modeler does not know the distributions but instead can only observe the choices and their outcomes for a user familiar with the decision problems, for example a skilled player playing a digital game. We propose a framework combining two separate components, one for modeling the unknown decision-making environment and another for the risk behavior. By using environment models capable of learning distributions we are able to infer classical models of decision-making under risk from observations of the user’s choices and outcomes alone, and we also demonstrate alternative models for predictive purposes. We validate the approach on artificial data and demonstrate a practical use case in modeling risk attitudes of professional esports teams.

The paper presents a novel model-based method for intelligent tutoring, with particular emphasis on the problem of selecting teaching interventions in interaction with humans. Whereas previous work has focused on either personalization of teaching or optimization of teaching intervention sequences, the proposed individualized model-based planning approach represents convergence of these two lines of research. Model-based planning picks the best interventions via interactive learning of a user memory model’s parameters. The approach is novel in its use of a cognitive model that can account for several key individual- and material-specific characteristics related to recall/forgetting, along with a planning technique that considers users’ practice schedules. Taking a rule-based approach as a baseline, the authors evaluated the method’s benefits in a controlled study of artificial teaching in second-language vocabulary learning (N = 53).