Geo-Spatial Location-based Visualization of Network Dynamics
By Nathan Sakunkoo and Patty Sakunkoo
Abstract
The state of the art of visualization has come a long way since the introduction of sociograms. However, with a few exceptions, much of the progress was based on the paradigm of visualizing static snapshots of nodes and edges, including the development of sophisticated layout algorithms. While insightful, these algorithms tend to disregard the physical locations of actors. Geo-spatial locations in real life can, nonetheless, augment perception, interpretation, and novel hypothesis development; yet, relatively little attention has been paid to incorporate this aspect into network visualization. Our work, netMap, is among the first attempts to explore the spatial dimension of the physical distance between actors (i.e. where they actually are located when the interactions occur), while also taking into account longitudinal data in producing dynamic visualization of changes in network structure. To facilitate the collection and display of physical locations, netMap offers two alternatives: (1) the researcher can choose to overlay an image (e.g. seating chart of students, organizational floor plan) as a background and then use netMap’s interactive interface to drag actors to locations in the map that correspond to their real-world locations. (2) netMap can automatically convert addresses (e.g. 19 Sunbelt St., San Diego) into geographical latitudes and longitudes and overlay the social actors’ locations on online maps such as YahooMap. netMap then visualizes dynamics of social interactions over time through a combination of network movies and semi-transparent images such as seating charts or maps overlaid as backgrounds. We believe that netMap’s location-based layout approach will offer new complementary insights into analysis of network dynamics. A video demonstrating netMap’s location-based layout visualization on top of SoNIA can be viewed at http://www.stanford.edu/~sakunkoo/netmap .