In this paper, we present our analysis of five expert interviews, each from a different application domain. Such analysis is crucial to understanding the real-world scenarios of analysing geographicallyembedded flow data. The results of our analysis show that similar high-level tasks were conducted in different domains. To better describe the targets of these tasks, we proposed three flow-targets for analysing geographically-embedded flow data: single flow, total flow and regional flow.
Index Terms:Human-centered computing-Visualization-Visualization design and evaluation methods
INTRODUCTIONMany domain experts and analysts around the world are interested in discovering insights and patterns related to some form of commodity flow between different geographic locations. Such activities include migration patterns, urban planning, animal movement, diesase mapping, financial trading and commuting behaviour.Geographically-embedded flow data, often termed spatial movement data, trajectory or origin-destination (OD) data, contains geographical locations (the origin, destination and potentially mid-way points), a connection or trajectory between them and a value referring to the magnitude of the flow between the locations. Data of this type is increasing with advances in technology, yet there are still limited tools to effectively analyse and visualise such data in order to discover patterns and inform decision making. In recent years we have seen growing support and solutions for the analysis of flow data, yet to the best of our knowledge, there is no research focusing on identifying real-world tasks conducted by the domain experts.This work was conducted as part of a 3.5 year PhD project focusing on exploring the design space of geographically-embedded flow visualisation [13, Chap. 2]. The motivation for these interviews was to understand the requirements and role of flow data in real-world applications and existing work flows across different disciplines, in order to help inform new visualisation designs [14,15].The contributions of this work are: 1) three high-level flow-targets for analysing geographically-embedded flow data; 2) using Munzner's framework [11] to abstract analytic tasks in interview analysis; 3) five use cases for motivating flow data visualisation design.
RELATED WORKTask Taxonomy -Visualisation tasks have been grouped and classified in a number of examples e.g. [2,9]. More recently, Brehmer and Munzner compiled these classifications into the what-why-how framework [6,11] to systematically abstract work flows for visual analytics. Specific taxonomies for geographical data visualisation has also been explored. Closest to this study is the study of Roth [12], who interviewed expert interactive map users to develop a taxonomy of interaction primitives for map-based visualisation. Earlier