Who, Where, When and with Whom? Evaluation of Group Meeting Visualizations
Visualizing time-dependent and location-based data is a challenging problem but highly relevant for areas like intelligence analysis, traffic control, or social network analysis. In this context, we address the problem of visualizing meetings between pers
- PDF / 1,002,446 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 14 Downloads / 163 Views
Vienna University of Technology, Vienna, Austria [email protected], {johanna.haider,margit}@igw.tuwien.ac.at 2 University of Applied Arts Vienna, Vienna, Austria [email protected]
Abstract. Visualizing time-dependent and location-based data is a challenging problem but highly relevant for areas like intelligence analysis, traffic control, or social network analysis. In this context, we address the problem of visualizing meetings between persons, groups of persons, vehicles, or other entities. However, the temporal dimension inherent in such data makes traditional map representations less well suited for this kind of problem as they easily become cluttered. To overcome this issue we developed a modified map representation and three alternative representations (two matrix-based visualizations and one based on Gantt charts). An empirical evaluation comparing these four visualizations and assessing correctness, recognition rates of groups, and subjective preference indicates that the alternative visualizations perform significantly better than the map-based representation when meetings need to be identified. In addition, we identify specific strengths and weaknesses of the investigated visualizations and propose design considerations. Keywords: Information visualization · Map · Matrix · Gantt charts
1 Introduction The analysis of spatio-temporal data to identify patterns of movements (e.g., movements of persons or groups of persons, vehicles, etc.) is of great interest in several domains such as location-based social networks analysis (e.g., [5, 15]), crime analysis (e.g., [4]), or movement pattern analysis (e.g., [3]). However, analysts and decisionmakers are often confronted with an enormous amount of multidimensional information that can be extracted from spatio-temporal data. In order to make these large amounts of data manageable there is a need to develop effective visualization methods. Especially in case of the analysis of group meetings not only questions relating to when and where are of interest but also more complex questions such as who meets with whom. For representing spatial data, map representations are very popular to show different types of geographic information in an intuitive way. Such maps can support analysts to deduce associations and connections from the spatio-temporal data (e.g., population density versus recreation areas) [17]. However, the temporal dimension is problematic c Springer International Publishing Switzerland 2016 M. Jamnik et al. (Eds.): Diagrams 2016, LNAI 9781, pp. 235–249, 2016. DOI: 10.1007/978-3-319-42333-3 19
236
S. Kriglstein et al.
with map representations as a single map usually only represents a single slice in time. Yet, this temporal aspect is vital for understanding meeting patterns of various people. Addressing this challenge, we designed alternative visualizations in the course of two projects to facilitate the analysis of location data of a group of individuals. It was important for us to find a possibility to answer where, when, w
Data Loading...
