David Reitter

A cognitive model of spatial path-planning

Planning a path to a destination, given a number of options and obstacles, is a common task. We suggest a two-component cognitive model that combines retrieval of knowledge about the environment with search guided by visual perception. In the first component, subsymbolic information, acquired during navigation, aids in the retrieval of declarative information representing possible paths to take. In the second component, visual information directs the search, which in turn creates knowledge for the first component. The model is implemented using the ACT-R cognitive architecture and makes realistic assumptions about memory access and shifts in visual attention. We present simulation results for memory-based high-level navigation in grid and tree structures, and visual navigation in mazes, varying relevant cognitive (retrieval noise and visual finsts) and environmental (maze and path size) parameters. The visual component is evaluated with data from a multi-robot control experiment, where subjects planned paths for robots to explore a building. We describe a method to compare trajectories without referring to aligned points in the itinerary. The evaluation shows that the model provides a good fit, but also that planning strategies may vary with task loads.

How groups develop a specialized domain vocabulary: A cognitive multi-agent model

We simulate the evolution of a domain vocabulary in small communities. Empirical data show that human communicators can evolve graphical languages quickly in a constrained task (Pictionary), and that communities converge towards a common language. We propose that simulations of such cultural evolution incorporate properties of human memory (cue-based retrieval, learning, decay). A cognitive model is described that encodes abstract concepts with small sets of concrete, related concepts (directing), and that also decodes such signs (matching). Learning captures conventionalized signs. Relatedness of concepts is characterized by a mixture of shared and individual knowledge, which we sample from a text corpus. Simulations show vocabulary convergence of agent communities of varied structure, but idiosyncrasy in vocabularies of each dyad of models. Convergence is weakened when agents do not alternate between encoding and decoding, predicting the necessity of bi-directional communication. Convergence is improved by explicit feedback about communicative success. We hypothesize that humans seek out subtle clues to gauge success in order to guide their vocabulary acquisition.

Exploration for Understanding in Cognitive Modeling

Exploration for Understanding in Cognitive Modeling The cognitive modeling and artificial general intelligence research communities may reap greater scientific return on research investments - may achieve an improved understanding of architectures and models - if there is more emphasis on systematic sensitivity and necessity analyses during model development, evaluation, and comparison. We demonstrate this methodological prescription with two of the models submitted for the Dynamic Stocks and Flows (DSF) Model Comparison Challenge, exploring the complex interactions among architectural mechanisms, knowledge-level strategy variants, and task conditions. To cope with the computational demands of these analyses we use a predictive analytics approach similar to regression trees, combined with parallelization on high performance computing clusters, to enable large scale, simultaneous search and exploration.

Metacognition and Multiple Strategies in a Cognitive Model of Online Control

Metacognition and Multiple Strategies in a Cognitive Model of Online Control We present a cognitive model performing the Dynamic Stocks&Flows control task, in which subjects control a system by counteracting a systematically changing external variable. The model uses a metacognitive layer that chooses a task strategy drawn from of two classes of strategies: precise calculation and imprecise estimation. The model, formulated within the ACT-R theory, monitors the success of each strategy continuously using instance-based learning and blended retrieval from declarative memory. The model underspecifies other portions of the task strategies, whose timing was determined as unbiased estimate from empirical data. The models predictions were evaluated on data collected from novel experimental conditions, which did not inform the models development and included discontinuous and noisy environmental change functions and a control delay. The model as well as the data show sudden changes in subject error and general learning of control; the model also correctly predicted oscillations of plausible magnitude. With its predictions, the model ranked first among the entries to the 2009 Dynamic Stocks&Flows modeling challenge.

Entropy Converges Between Dialogue Participants: Explanations from an Information-Theoretic Perspective.

The applicability of entropy rate constancy to dialogue is examined on two spoken dialogue corpora. The principle is found to hold; however, new entropy change patterns within the topic episodes of dialogue are described, which are different from written text. Speaker’s dynamic roles as topic initiators and topic responders are associated with decreasing and increasing entropy, respectively, which results in local convergence between these speakers in each topic episode. This implies that the sentence entropy in dialogue is conditioned on different contexts determined by the speaker’s roles. Explanations from the perspectives of grounding theory and interactive alignment are discussed, resulting in a novel, unified informationtheoretic approach of dialogue.

Spectral Analysis of Information Density in Dialogue Predicts Collaborative Task Performance

We propose a perspective on dialogue that focuses on relative information contributions of conversation partners as a key to successful communication. We predict the success of collaborative task in English and Danish corpora of task-oriented dialogue. Two features are extracted from the frequency domain representations of the lexical entropy series of each interlocutor, power spectrum overlap (PSO) and relative phase (RP). We find that PSO is a negative predictor of task success, while RP is a positive one. An SVM with these features significantly improved on previous task success prediction models. Our findings suggest that the strategic distribution of information density between interlocutors is relevant to task success.

A cognitive model of perceptual path planning in a multi-robot control system

We discuss an experiment involving visual path planning for multiple, remote robots in a partially visible building, with a partial 2D map available. Participants in the experiment defined waypoints for each robot to circumnavigate obstacles and explore the building. A cognitively plausible model of visual planning is evaluated using a normalized metric of the fit between model and subject itineraries. We discuss variation in the data and model fit, indicating individual differences in strategies to cope with task demands.