In the past several decades, considerable theoretical progress has been made in understanding the role of reference frames in the encoding and retrieval of spatial information about the environment. Many of these insights have come from participants making judgments of relative direction using their memories of spatial layouts. In this task, participants are asked to imagine standing at a given location and facing a certain direction, and to point to a target location. Although this task has been widely and productively used, a computational cognitive model of judgments of relative direction has yet to be introduced. Computational modeling of judgments of relative direction is a critical next step to formulating and testing hypotheses about the cognitive processes involved in establishing and using spatial reference frames. We present an initial attempt to model judgments of relative direction and fit the model to two datasets exhibiting behavioral patterns commonly observed in the spatial memory literature. The model was able to predict many important features of these data, most notably alignment effects. We discuss directions for future modeling efforts.

在过去的几十年中，在理解参考框架在编码和检索环境空间信息中的作用方面取得了相当大的理论进展。其中许多见解来自参与者使用他们对空间布局的记忆来判断相对方向。在这项任务中，要求参与者想象站在给定位置并面向某个方向，并指向目标位置。尽管这项任务已被广泛且富有成效地使用，但尚未引入相对方向判断的计算认知模型。相对方向判断的计算建模是制定和测试关于建立和使用空间参考框架所涉及的认知过程的假设的关键下一步。我们提出了对相对方向判断进行建模的初步尝试，并将模型拟合到两个数据集，这些数据集展示了空间记忆文献中常见的行为模式。该模型能够预测这些数据的许多重要特征，最显着的是对齐效应。我们讨论未来建模工作的方向。