posted on 2004-08-01, 00:00authored byGlenn Gunzelmann, John R. Anderson
This paper presents a model of human spatial orientation, using a task that involves locating targets on a map of the space. The model uses a hierarchical solution process that was reported by many of the participants in an empirical study. It encodes the location of the target in the visual scene by identifying a cluster of objects that contains the target and then encoding the position of the target within that cluster. By applying this description of the target’s location to the representation on the map of the space, the model is able to correctly identify the target. Using this general strategy, it reproduces all of the major trends in the empirical data.