Under Review

Barrett, J. A. and VanDrunen, J. (2021). “Language Games and the Emergence of Discourse.” [Preprint]

Ludwig Wittgenstein (1958) used the notion of a language game to illustrate how language is interwoven with action. Here we consider how successful linguistic discourse of the sort he described might emerge in the context of a self-assembling evolutionary game. More specifically, we consider how discourse and coordinated action might self-assemble in the context of two generalized signaling games. The first game shows how prospective language users might learn to initiate meaningful discourse. The second shows how more subtle varieties of discourse might co-emerge with a meaningful language.


VanDrunen, J. and Pizlo, Z. (2019). “The Effectiveness of Multidimensional Scaling in TSPs Whose Metric is not Euclidean.” Society for Mathematical Psychology (Montréal, July 2019). [Poster]

It is commonly assumed that 2D images are represented as 2D Euclidean planes in the human visual system. This assumption has received support from numerous studies in which human subjects produced near-optimal Traveling Salesman (TSP) tours. Specifically, the human subjects produced TSP tours in a sequence of coarse-to-fine approximations by using a hierarchical clustering (pyramid) representation of the problem. When obstacles are introduced into a 2D Euclidean TSP, the distances between vertices are no longer Euclidean, but human subjects can still produce near-optimal tours, as long as the obstacles are geometrically simple (line segments, L, and C shapes). Can this pyramid algorithm be applied to a Euclidean approximation of the pairwise distances produced by Multidimensional Scaling (MDS), which is often the method of choice for representing cognitive spaces? In this project, we evaluated the usefulness of MDS in visual tasks in which the ground truth of the geometrical distances is known. We did this by applying a graph-pyramid algorithm, as well as the optimal Concorde algorithm to Euclidean approximations of a TSP (i.e., a TSP without obstacles) produced by MDS when MDS was applied to a TSP with obstacles. Straight line segments were used as obstacles. Errors of the pyramid algorithm and the Concorde algorithm with MDS approximations were measured in 2D, 3D and 4D spaces. The TSP solutions produced by the pyramid and the Concorde algorithm show substantial improvement between the 2D and 3D Euclidean approximations produced by MDS, but this improvement was only observed with larger TSPs.


Buibas, M., Quinn, J., Bapst, A., Khorsandi, R., Shah, N., VanDrunen, J., Wildie, M., and Yousefisahi, S (pending). “System that Performs Selective Manual Review of Shopping Carts in an Automated Store.” [Google Patents]

An automated store that calculates a confidence score for virtual shopping carts of shoppers, and selects carts for manual review based on these scores. Carts with low confidence scores may be more likely to contain errors, so prioritizing manual review of these carts is a cost-effective method of improving overall accuracy. A cart confidence score may be a function of factors such as confidence in the trajectory of the shopper generated by the store tracking system, confidence in the events (such as taking an item from a shelf) that affect the cart, and confidence that events are attributed to the correct shopper. Situations that make tracking, item identification, or attribution more complex may reduce confidence levels. For example, attribution confidence may be low when multiple shoppers are near an event, and item confidence may be low if the probabilistic classifier that identifies the item assigns nontrivial probabilities to multiple items.