It is proposed that there are important generalizations about problem solving in design activity that reach across specific disciplines. A framework for the study of design is presented that (a) characterizes design as a radial category and fleshes out the task environment of the prototypical cases; (b) takes the task environment seriously; (c) shows that this task environment occurs in design tasks, but does not occur in every nondesign task; (d) explicates the impact of this task environment on the design (...) problem space; and (e) demonstrates that, given the structure of the information-processing system, the features noted in the problem spaces of design tasks will not all occur in problem spaces where the task environment is vastly different. This analysis leads to the claim that there are a set of invariant features in the problem spaces of design situations that collectively constitute a design problem space. Protocol studies are reported in which the problem spaces of three design tasks in architecture, mechanical engineering, and instructional design are explored and compared with several protocols from nondesign problem-solving tasks. (shrink)

This article describes rational analyses and cognitive models of Web users developed within information foraging theory. This is done by following the rational analysis methodology of (a) characterizing the problems posed by the environment, (b) developing rational analyses of behavioral solutions to those problems, and (c) developing cognitive models that approach the realization of those solutions. Navigation choice is modeled as a random utility model that uses spreading activation mechanisms that link proximal cues (information scent) that occur in Web browsers (...) to internal user goals. Web‐site leaving is modeled as an ongoing assessment by the Web user of the expected benefits of continuing at a Web site as opposed to going elsewhere. These cost–benefit assessments are also based on spreading activation models of information scent. Evaluations include a computational model of Web user behavior called Scent‐Based Navigation and Information Foraging in the ACT Architecture, and the Law of Surfing, which characterizes the empirical distribution of the length of paths of visitors at a Web site. (shrink)

Science is a form of distributed analysis involving both individual work that produces new knowledge and collaborative work to exchange information with the larger community. There are many particular ways in which individual and community can interact in science, and it is difficult to assess how efficient these are, and what the best way might be to support them. This paper reports on a series of experiments in this area and a prototype implementation using a research platform called CACHE. CACHE (...) both supports experimentation with different structures of interaction between individual and community cognition and serves as a prototype for computational support for those structures. We particularly focus on CACHE‐BC, the Bayes community version of CACHE, within which the community can break up analytical tasks into “mind‐sized” units and use provenance tracking to keep track of the relationship between these units. (shrink)

Some of the required characteristics for a true machine theory of mind (MToM) include the ability to (1) reproduce the full diversity of human thought and behavior, (2) develop a personalized model of an individual with very limited data, and (3) provide an explanation for behavioral predictions grounded in the cognitive processes of the individual. We propose that a certain class of cognitive models provide an approach that is well suited to meeting those requirements. Being grounded in a mechanistic framework (...) like a cognitive architecture such as ACT-R naturally fulfills the third requirement by mapping behavior to cognitive mechanisms. Exploiting a modeling paradigm such as instance-based learning accounts for the first requirement by reflecting variations in individual experience into a diversity of behavior. Mechanisms such as knowledge tracing and model tracing allow a specific run of the cognitive model to be aligned with a given individual behavior trace, fulfilling the second requirement. We illustrate these principles with a cognitive model of decision-making in a search and rescue task in the Minecraft simulation environment. We demonstrate that cognitive models personalized to individual human players can provide the MToM capability to optimize artificial intelligence agents by diagnosing the underlying causes of observed human behavior, projecting the future effects of potential interventions, and managing the adaptive process of shaping human behavior. Examples of the inputs provided by such analytic cognitive agents include predictions of cognitive load, probability of error, estimates of player self-efficacy, and trust calibration. Finally, we discuss implications for future research and applications to collective human–machine intelligence. (shrink)

We argue that cognitive models can provide a common ground between human users and deep reinforcement learning (Deep RL) algorithms for purposes of explainable artificial intelligence (AI). Casting both the human and learner as cognitive models provides common mechanisms to compare and understand their underlying decision-making processes. This common grounding allows us to identify divergences and explain the learner's behavior in human understandable terms. We present novel salience techniques that highlight the most relevant features in each model's decision-making, as well (...) as examples of this technique in common training environments such as Starcraft II and an OpenAI gridworld. (shrink)

Animals routinely adapt to changes in the environment in order to survive. Though reinforcement learning may play a role in such adaptation, it is not clear that it is the only mechanism involved, as it is not well suited to producing rapid, relatively immediate changes in strategies in response to environmental changes. This research proposes that counterfactual reasoning might be an additional mechanism that facilitates change detection. An experiment is conducted in which a task state changes over time and the (...) participants had to detect the changes in order to perform well and gain monetary rewards. A cognitive model is constructed that incorporates reinforcement learning with counterfactual reasoning to help quickly adjust the utility of task strategies in response to changes. The results show that the model can accurately explain human data and that counterfactual reasoning is key to reproducing the various effects observed in this change detection paradigm. (shrink)