The aim of the contribution is to present results of eye-tracking experiments on evaluation of user preferences during reading 2D and 3D cartographic visualizations. Currently, thanks to advances in computer and Internet technology, the production of digital cartographic products is massive. Map makers, cartographers or graphic designers perceive maps differently than the target audience. The considerable degree of subjectivity is put into during the map making. In many cases, the specialist cannot imagine how the map will be used. For these reasons, it is necessary to make research of user perception and cognition of maps.
In modern cartography, it is very popular to depict the spatial information using 3D visualization techniques, perspective views and pseudo - 3D techniques like hillshade, hill or hatch hypsometry methods. The research question is to find out the real value of 3D cartographic methods for the perception and use of maps. 3D maps are generally considered as a way how to better show the vertical spatial relations, while classical 2D representations (orthogonal maps) are regarded as more suitable for distance and area perception. Both mentioned visualization methods has pros and cons, and it is necessary to objectively specify, which one is suitable for solving different spatial tasks.
To evaluate the utilization of 3D and 2D visualization methods, two eye-tracking experiments were performed using the SMI RED 250 eye-tracker with 120 Hz sample rate. For experiment design, the manufacturer software Experiment Center was used. The data analyses were executed in BeGaze, OGAMA and R software. The first experiment was designed as a set of stimuli containing single maps. Half of them was using 3D visualization, whereas the second half was in 2D. Purpose of this experiment was to evaluate users behaviour during answering the spatial query (e.g. Find the highest peak, Find the furthest point, etc.). The second experiment was focused on finding out the user preferences between both visualization methods. Stimuli was represented as a pair of maps in 2D and 3D side by side.
Results of the experiments are based on statistical analyses of various eye-tracking metrics (fixation duration, fixation/saccades ratio, AOI dwell time, time on task). The overall goal of the research is to create a theoretical framework for investigating effectiveness and preferences of 3D and 2D cartographic visualizations. Eye-tracking technology was not fully utilized in the cartography or geosciences yet. It is clear that it will have great importance in optimization of cartographic products and visualization of geographic data in the future.