USER TYPOLOGY BASED ON EYE-MOVEMENT PATHS

References: 13th SGEM GeoConference on Informatics, Geoinformatics And Remote Sensing, www.sgem.org, SGEM2013 Conference Proceedings, ISBN 978-954-91818-9-0 / ISSN 1314-2704, June 16-22, 2013, Vol. 1, 1041 - 1048 pp

ABSTRACT

Maps can be read in many different ways. Cartographers look at maps differently from non-cartographers. There should be a difference in map cognition between males and females, between young and old people, and in accordance with many other characteristics.
The paper focuses on map readers’ typology based on their eye-movements during solving the geographical tasks with use of a map. Within the study, a device for recording eye-movements – an eye-tracker - was used. Eye movements can be analyzed using different methods. For user typology, most appropriate should be the ScanPath comparison. This approach was previously used by Fabrikant (2008) and Coltekin (2009).
The first method, introduced by Privitera and Stark (2000) uses the comparison based on string editing. The fixations are replaced with characters standing for the Areas of Interest (AOIs) they hit and the ScanPath is represented as a string of characters. This approach was used in the analyses of stimuli, in which the Areas of Interest were clearly defined. For data analyses and visualization, eyePatterns software was used.
The second method is called Path Similarity Analysis (PSA) and was introduced by Andrienko et al. (2012). This method directly compares the paths resulting from the raw data (in contrast to the previous method, which compares only sequences of visited AOIs).
Both methods produce clusters of users, based on the movement of their eyes. The methods were tested and their results compared to each other. Clusters were compared to data gathered from respondents (age, sex, cartographical education) and it was found that some of these clusters correspond to these data. For that reason it is possible to say that specific groups of respondents appear in the map in specific way.
The contribution describes the design of the eye-tracking experimental study and the process of data evaluation using both the aforementioned methods of ScanPath comparison.

Keywords: Cartography, Eye-tracking, ScanPath comparison, Typology, Cognition