SGER - Studying Map Interaction Behavior

University Of North Carolina At Charlotte, Charlotte NC

Investigators

Abstract

Geographic maps are one of the most traditional, familiar, and important visualizations in widespread use today. There is increasing understanding of the importance of visualization in providing insight into a vast amount of data to support critical reasoning and decision making in a variety of domains. Moreover, in the past few years, there has been a revolution in online access to map data and basic integration tools, enabling user driven applications from remote archaeological research to "mashups" mapping avian flu outbreak, military testing, local gas prices and far more. This represents a fundamental shift from few general-use map visualizations developed by expert cartographers to many specific map visualizations being developed by end users. As the latter utilize a growing number of mapping tools, new interaction methods and information management approaches are needed to reduce visualization complexity and improve usability. Yet research into map use has focused primarily on the artifact of visual representation and very little on the dynamic behavior of users' interactions with those representations. There are few guidelines for creating map visualizations that meet user needs, potentially limiting the effectiveness of these critical visualizations. Thus, there is a recognized need for a deep investigation into how people use maps to support the growing efforts in visualization and visual analytics. This proposal addresses that need by using Human Computer Interaction techniques in to study the usability of map visualizations. Specifically, the project will address the following four questions: 1.) How do users interact with map visualizations? 2.) How do users deal with map complexity? 3.) How do elements of the map visualizations support user behavior? 4.) How do interactions vary across user and task context? Specifically, the project will consist of two exploratory, observational studies of users performing a variety of map-based analysis tasks, examining their interaction behavior with map visualizations. Outcomes will identify the usability problems and issues that users face, as well as the common patterns of interaction for particular tasks and more generally across tasks. These initial results will lay the groundwork for theories on the use and design of interactive map visualizations, advancing our understanding of how users interpret geovisualizations. Because the problem has received very little scholarly attention, this project will be a foray into largely untested ideas. Further, given the recent explosion of interactive, online, and user-centered mapping, project outcomes have significant potential to establish transformative directions for geovisualization and human-centered computing. Outcomes will also ground work on automatically adapting and personalizing geovisualizations based on user and task contexts. Broader Impacts Geovisualization is a powerful mechanism to enable critical thinking and decision making involving spatial information, and geographic maps are the most widely used geovisualization. Understanding how people use and interact with maps will guide the improved design of these critical visualizations, potentially impacting a wide variety of domains, such as public health, environmental science, and emergency planning. This project represents a first-of-its-kind evaluation of geovisualizations from an interaction behavior point of view. Our preliminary results will provide design guidelines for interactive maps and demonstrate the value of interaction research in geovisualization.