Wayfinding Choremes 2.0 – Conceptual Primitives as a Basis for Translating Natural into Formal Language

Continuous spatial and temporal information is cognitively segmented into meaningful units that we refer to here as conceptual primitives; since we are focusing on routes in street networks, we define these primitives more specifically as wayfinding choremes. For the case of movement patterns in constraining environments such as driving around in street networks the wayfinding choremes are coupled with decision points that can offer alternative path choices or structure and/ or confirm a specific part of the route. To be able to analyze these choremes we have conducted a series of experiments in the past that focused on how a movement pattern can be characterized linguistically. Here, we apply the insights we gained to analyze and interpret linguistically characterized movement patterns to questions of automatically identifying movement patterns in text. That means we are analyzing verbal route directions and identify wayfinding choremes to infer the underlying movement patterns. To support our long term goal of automatic extraction of wayfinding choremes from text documents, we formally specify a taxonomy of wayfinding choremes as a first step for an ontological analysis. We will present the theoretical underpinning, tools, first results, and our perspectives on the linguistic analysis of movement patterns in this contribution. We hope to contribute to a better understanding of how natural language descriptions and formal characterizations can be closer related and can inform the automatic extraction of movement patterns from text.