Markus Kattenbeck

Empirically measuring salience of objects for use in pedestrian navigation

In this paper a Structural Equation Model drawing on current theories of salience is empirically tested using a large scale in-situ experiment (no = 366 objects and np = 119 participants). Using estimation methods based on partial least squares strong empirical evidence is found for the ability of the model to predict salience. 72% of the variance present in overall salience can be explained. Formative measurement of visual salience is revealed to be an appropriate way to measure visual salience, as the convergent validity analysis yields a highly significant path coefficient of 0.810. Route related features and visual aspects turn out to be most and equally important to predict overall salience, whereas rather person-related dimensions turn out to be less important. Overall, the model presented provides a reasonable and empirically sound way of measuring salience of objects in a survey-based manner.

Do You Recognize That Building’s Façade?

With the computational power of modern smartphones constantly increasing, resource intensive applications are becoming feasible to an ever growing extent. In this paper, we report on a research project recently started. Its aim is to develop an application for smartphones that combines pedestrian and public transport navigation including the computation of routes consisting of pedestrian routes and public transport trips and intuitive user guidance at any time of the trip. In particular, we focus on intuitive user guidance based on (LMs) in the surroundings of the user. For this reason, we use collaborative approaches to collect LMs and data about them.

Towards interfaces of mobile pedestrian navigation systems adapted to the user’s orientation skills

Abstract It is an open question, whether differently developed orientation skills require alternate pedestrian navigation interface designs for optimal self-localization during wayfinding. To address this issue we conducted two user studies in order to analyze different presentations of mobile maps. Evaluations were conducted in an outdoor ( N = 112 ) and indoor ( N = 64 ) environment. We recorded the time it took participants to localize themselves with the help of salient objects. Significant results were found for both studies. Abstract interfaces support well-oriented users, whereas worse oriented ones profited from standard designs. Consequently, we make the case for user-adaptive pedestrian navigation interfaces.

Estimating Models Combining Latent and Measured Variables: A Tutorial on Basics, Applications and Current Developments in Structural Equation Models and their Estimation using PLS Path Modeling

Structural Equation Modeling is a powerful statistical approach where measured variables and those which are latent can be combined in a single model. In this half-day tutorial participants learned about the statistical technique, its theoretical underpinnings and gained sufficient insight to apply this technique in a practical sense to their own research problems.

How Subdimensions of Salience Influence Each Other. Comparing Models Based on Empirical Data.

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.

Manipulating the Perception of Credibility in Refugee Related Social Media Posts

This paper describes a controlled web-based study (n=126), investigating whether the perception of the credibility of refugee-related Tweets can be influenced by cues already reported in the literature for social media content generally. We provide empirical evidence that both a Tweets popularity and the presence of links -- even neutral links created by URL shortening services -- may increase a users belief that the Tweet contains credible information. This is important because the propagation of false information relating to refugees on social media sites has been well documented.

Der Weg ist das Ziel – Fußgängernavigation ist Forschung zu Information Behavior

Aktuelle Forschungsarbeiten zur Fußgängernavigation, die im Rahmen des Forschungsprojekts NADINE am Lehrstuhl für Informationswissenschaft durchgeführt werden, profitieren stark von den Facetten Information Seeking, Context und Information Avoiding der Information Behavior-Forschung. Dabei wird zunächst von völlig technologieunabhängigen Erkenntnissen zu den Verhaltensweisen im Kontext der Wegfindung ausgegangen. Darauf aufbauend wird beispielhaft die Schaffung der nötigen Datengrundlage und die experimentelle Untersuchung von navigationsunterstützenden Informationssystemen im Innenbereich näher betrachtet. Information Seeking, Context and Information Avoiding are well-known concepts of Information Behavior research. All of these are of major importance for research work on pedestrian navigation systems (which is part of the research project NADINE). Consequently, the work presented in this article sets out on findings of wayfinding behavior. Indoor navigation is used as an example to illustrate two major steps involved in pedestrian navigation system research: building the basis for route calculation and empirically investigate information systems, which facilitate human wayfinding. Les projets de recherche actuels sur la navigation piétonne, qui sont effectuées dans le cadre du projet NADINE au Département de sciences de l’information, profitent largement des facettes « Information Seeking » (recherche d’information), « Context » (contexte) et « Information Avoiding » (évitement de l’information) de la recherche sur le comportement en matière d’information. Les auteurs sont partis d’abord de connaissances des pratiques dans le contexte de la navigation, qui sont complètement indépendantes de la technologie. En se basant sur ces connaissances, ils ont examiné de plus près la création des données de base nécessaires et l’enquête expérimentale des systèmes d’information, soutenus par la navigation, pour une utilisation en intérieur.

Do Spatial Abilities Have an Impact on Route Learning in Hypertexts

Metaphors of navigation have been widely used to describe the behaviour of users surfing the World Wide Web. We present the results of a web-based experiment (\(N=85\) participants) on route learning in Wikipedia. As spatial abilities and sense of direction are known to be important for real-world wayfinding abilities, we examine the extent to which the participants are able to retrace a learned route on their own and the time taken to do this can be predicted using these variables. The tested (G)LMM models, however, show a lower than expected relevance of spatial abilities and sense of direction. The results suggest that both personal factors (such as age and gender) and task are important for the duration of tasks.

Dissertation Abstract: Empirically Measuring Salience of Objects for Use in Pedestrian Navigation

What is still missing (cf. [2, 8]) is a large scale experiment, bridging theories and empirical evidence. It is of benefit to empirically reveal which of the known dimensions are most important and how the different sub-dimensions of salience relate to each other because this kind of knowledge can foster the development of theories about the salience of landmarks as well as its empirical measurement. The thesis abstracted here has three aims, consequently:

ISI 2015 Conference Review

The International Symposium of Information Science 2015 celebrated the 25th anniversary of the conference series with fruitful discussions in a lively atmosphere in Zadar, Croatia. The conference review summarizes the keynotes by Tefko Saracevic and Julianne Nyhan as well as Lyn Robinson’s talk about Information Science and Digital Humanities and a selection of talks given at the pre-conference workshop “Digital humanities and the technologies of the semantic web: decolonizing description for the sake of digital humanities”.