Amultidimensional model for selecting personalised landmarks

摘要

Route instructions for pedestrians usually include many references to landmarks. The specific landmark used depends on the perceived familiarity of the traveller with the environment. While a human informant is able to perceive this personal dimension and react accordingly, existing landmark selection algorithms neglect the individual spatial knowledge of a wayfinder. In this paper, we share our ideas for incorporating a personal dimension into the definition of landmark salience. This complements current models, which identify landmark salience based on attributes of landmarks or integrate route-dependent landmarks based on environmental factors. We propose a conceptual framework for a multidimensional model for personalised landmarks that integrates three dimensions: a dimension describing the landmark, an environmental dimension and a personal dimension, which has been lacking in existing landmark salience models. We identify and discuss attributes and attribute values for each of the dimensions. Furthermore, salience measures for the attributes of the individual dimensions are developed. The novelty of this research consists of integrating all known attributes and potential values within a single model of salience, while focusing on the personal dimension.