文章基本信息

标题：How Subdimensions of Salience Influence Each Other. Comparing Models Based on Empirical Data
本地全文：下载
作者：Markus Kattenbeck
期刊名称：LIPIcs : Leibniz International Proceedings in Informatics
电子版ISSN：1868-8969
出版年度：2017
卷号：86
页码：10:1-10:13
DOI：10.4230/LIPIcs.COSIT.2017.10
出版社：Schloss Dagstuhl -- Leibniz-Zentrum fuer Informatik
摘要：Theories about salience of landmarks in GIScience have been evolving for about 15 years. This paper empirically analyses hypotheses about the way different subdimensions (visual, structural, and cognitive aspects, as well as prototypicality and visibility in advance) of salience have an impact on each other. The analysis is based on empirical data acquired by means of an in-situ survey (360 objects, 112 participants). It consists of two parts: First, a theory-based structural model is assessed using variance-based Structural Equation Modeling. The results achieved are, second, corroborated by a data-driven approach, i.e. a tree-augmented naive Bayesian network is learned. This network is used as a structural model input for further analyses. The results clearly indicate that the subdimensions of salience influence each other.
关键词：Salience models; consistent PLS-SEM Analysis; Bayesian Networks