Linking cognitive and computational saliences in route information

Abstract

Finding a destination in a new spatial environment can be a daunting task. To aid navigation, many people take advantage of route directions, either provided by other people or by electronic navigation services. However, their effectiveness may be hampered if they are overly complex. While most people are generally good at focusing on important information, this is a challenge for navigation services. Thus, being able to automatically determine important points along a route that need to be included in route directions would provide a further step towards cognitively ergonomic navigation services. In the present study, methods for calculating the salience-or importance-of decision points are correlated with the frequency of decision points appearing in route directions. Results show that metrics based on the probability of a decision point being traversed and information-theoretic quantities of decision points correlate significantly with incidence in route directions, indicating that it is possible to identify crucial decision points in advance. This has implications for the design of navigation services that are able to adapt their assistance in real time. © 2012 Springer-Verlag.