An empirically-derived taxonomy of interaction primitives for interactive cartography and geovisualization.

Proposals to establish a 'science of interaction' have been forwarded from Information Visualization and Visual Analytics, as well as Cartography, Geovisualization, and GIScience. This paper reports on two studies to contribute to this call for an interaction science, with the goal of developing a functional taxonomy of interaction primitives for map-based visualization. A semi-structured interview study first was conducted with 21 expert interactive map users to understand the way in which map-based visualizations currently are employed. The interviews were transcribed and coded to identify statements representative of either the task the user wished to accomplish (i.e., objective primitives) or the interactive functionality included in the visualization to achieve this task (i.e., operator primitives). A card sorting study then was conducted with 15 expert interactive map designers to organize these example statements into logical structures based on their experience translating client requests into interaction designs. Example statements were supplemented with primitive definitions in the literature and were separated into two sorting exercises: objectives and operators. The objective sort suggested five objectives that increase in cognitive sophistication (identify, compare, rank, associate, & delineate), but exhibited a large amount of variation across participants due to consideration of broader user goals (procure, predict, & prescribe) and interaction operands (space-alone, attributes-in-space, & space-in-time; elementary & general). The operator sort suggested five enabling operators (import, export, save, edit, & annotate) and twelve work operators (reexpress, arrange, sequence, resymbolize, overlay, pan, zoom, reproject, search, filter, retrieve, & calculate). This taxonomy offers an empirically-derived and ecologically-valid structure to inform future research and design on interaction.

Designing a web-based learning portal for geographic visualization and analysis in public health.

Interactive mapping and spatial analysis tools are under-utilized by health researchers and decision-makers as a result of scarce training materials, few examples demonstrating the successful use of geographic visualization, and poor mechanisms for sharing results generated by geovisualization. Here, we report on the development of the Geovisual EXplication(G-EX) Portal, a web-based application designed to connect researchers in geovisualization and related mapping sciences, to users who are working in public health and epidemiology. This paper focuses on the design and development of the G-EX Portal Learn module, a set of tools intended to disseminate learning artifacts. Initial design and development of the G-EX Portal has been guided by our past research on the use and usability of geovisualization in public health. As part of the iterative design and development process, we conducted a needs assessment survey with targeted end-users, which we report on here. The survey focused on users' current learning habits, their preferred kind of learning artifacts and issues they may have with contributing learning artifacts to web portals. Survey results showed that users desire a diverse set of learning artifacts in terms of both formats and topics covered. Results also revealed a willingness of users to contribute both learning artifacts and personal information that would help other users to evaluate the credibility of the learning artifact source. We include a detailed description of the G-EX Portal Learn module and focus on modifications to the design of the Learn module as a result from feedback we received from our survey.

The roles of social domains, behavioral risk, health care resources, and chlamydia in spatial clusters of US cervical cancer mortality: not all the clusters are the same.

BACKGROUND: While high-risk geographic clusters of cervical cancer mortality have previously been assessed, factors associated with this geographic patterning have not been well studied. Once these factors are identified, etiologic hypotheses and targeted population-based interventions may be developed and lead to a reduction in geographic disparities in cervical cancer mortality. METHODS: The authors linked multiple data sets at the county level to assess the effects of social domains, behavioral risk factors, local physician and hospital availability, and Chlamydia trachomatis infection on overall spatial clustering and on individual clusters of cervical cancer mortality rates in 2000-2004 among 3,105 US counties in the 48 states and the District of Columbia. RESULTS: During the study period, a total of 19,898 cervical cancer deaths occurred in women aged 20 and older. The distributions of county-level characteristics indicated wide ranges in social domains measured by demographics and socioeconomic status, local health care resources, and the rate of chlamydial infection. We found that overall geographic clustering of increased cervical cancer mortality was related to the high proportion of black population, low socioeconomic status, low Papanicolaou test rate, low health care coverage, and the high chlamydia rate; however, unique characteristics existed for each individual cluster, and the Appalachian cluster was not related to a high proportion of black population or to chlamydia rates. DISCUSSION: This study indicates that local social domains, behavioral risk, and health care sources are associated with geographic disparities in cervical cancer mortality rates. The association between the chlamydia rate and the cervical cancer mortality rate may be confounded by other factors known to be a risk for cervical cancer mortality, such as the infection with human papillomavirus. The findings will help cancer researchers examine etiologic hypotheses and develop tailored, cluster-specific interventions to reduce cervical cancer disparities.

Value-by-alpha maps: An alternative technique to the cartogram.

The cartogram, or value-by-area map, is a popular technique for cartographically representing social data. Such maps visually equalize a basemap prior to mapping a social variable by adjusting the size of each enumeration unit by a second, related variable. However, to scale the basemap units according to an equalizing variable, cartograms must distort the shape and/or topology of the original geography. Such compromises reduce the effectiveness of the visualization for elemental and general map-reading tasks. Here we describe a new kind of representation, termed a value-by-alpha map, which visually equalizes the basemap by adjusting the alpha channel, rather than the size, of each enumeration unit. Although not without its own limitations, the value-by-alpha map is able to circumvent the compromise inherent to the cartogram form, perfectly equalizing the basemap while preserving both shape and topology.

A workflow learning model to improve geovisual analytics utility.

INTRODUCTION: This paper describes the design and implementation of the G-EX Portal Learn Module, a web-based, geocollaborative application for organizing and distributing digital learning artifacts. G-EX falls into the broader context of geovisual analytics, a new research area with the goal of supporting visually-mediated reasoning about large, multivariate, spatiotemporal information. Because this information is unprecedented in amount and complexity, GIScientists are tasked with the development of new tools and techniques to make sense of it. Our research addresses the challenge of implementing these geovisual analytics tools and techniques in a useful manner. OBJECTIVES: The objective of this paper is to develop and implement a method for improving the utility of geovisual analytics software. The success of software is measured by its usability (i.e., how easy the software is to use?) and utility (i.e., how useful the software is). The usability and utility of software can be improved by refining the software, increasing user knowledge about the software, or both. It is difficult to achieve transparent usability (i.e., software that is immediately usable without training) of geovisual analytics software because of the inherent complexity of the included tools and techniques. In these situations, improving user knowledge about the software through the provision of learning artifacts is as important, if not more so, than iterative refinement of the software itself. Therefore, our approach to improving utility is focused on educating the user. METHODOLOGY: The research reported here was completed in two steps. First, we developed a model for learning about geovisual analytics software. Many existing digital learning models assist only with use of the software to complete a specific task and provide limited assistance with its actual application. To move beyond task-oriented learning about software use, we propose a process-oriented approach to learning based on the concept of scientific workflows. Second, we implemented an interface in the G-EX Portal Learn Module to demonstrate the workflow learning model. The workflow interface allows users to drag learning artifacts uploaded to the G-EX Portal onto a central whiteboard and then annotate the workflow using text and drawing tools. Once completed, users can visit the assembled workflow to get an idea of the kind, number, and scale of analysis steps, view individual learning artifacts associated with each node in the workflow, and ask questions about the overall workflow or individual learning artifacts through the associated forums. An example learning workflow in the domain of epidemiology is provided to demonstrate the effectiveness of the approach. RESULTS/CONCLUSIONS: In the context of geovisual analytics, GIScientists are not only responsible for developing software to facilitate visually-mediated reasoning about large and complex spatiotemporal information, but also for ensuring that this software works. The workflow learning model discussed in this paper and demonstrated in the G-EX Portal Learn Module is one approach to improving the utility of geovisual analytics software. While development of the G-EX Portal Learn Module is ongoing, we expect to release the G-EX Portal Learn Module by Summer 2009.

Geovisual analytics to enhance spatial scan statistic interpretation: an analysis of U.S. cervical cancer mortality.

BACKGROUND: Kulldorff's spatial scan statistic and its software implementation - SaTScan - are widely used for detecting and evaluating geographic clusters. However, two issues make using the method and interpreting its results non-trivial: (1) the method lacks cartographic support for understanding the clusters in geographic context and (2) results from the method are sensitive to parameter choices related to cluster scaling (abbreviated as scaling parameters), but the system provides no direct support for making these choices. We employ both established and novel geovisual analytics methods to address these issues and to enhance the interpretation of SaTScan results. We demonstrate our geovisual analytics approach in a case study analysis of cervical cancer mortality in the U.S. RESULTS: We address the first issue by providing an interactive visual interface to support the interpretation of SaTScan results. Our research to address the second issue prompted a broader discussion about the sensitivity of SaTScan results to parameter choices. Sensitivity has two components: (1) the method can identify clusters that, while being statistically significant, have heterogeneous contents comprised of both high-risk and low-risk locations and (2) the method can identify clusters that are unstable in location and size as the spatial scan scaling parameter is varied. To investigate cluster result stability, we conducted multiple SaTScan runs with systematically selected parameters. The results, when scanning a large spatial dataset (e.g., U.S. data aggregated by county), demonstrate that no single spatial scan scaling value is known to be optimal to identify clusters that exist at different scales; instead, multiple scans that vary the parameters are necessary. We introduce a novel method of measuring and visualizing reliability that facilitates identification of homogeneous clusters that are stable across analysis scales. Finally, we propose a logical approach to proceed through the analysis of SaTScan results. CONCLUSION: The geovisual analytics approach described in this manuscript facilitates the interpretation of spatial cluster detection methods by providing cartographic representation of SaTScan results and by providing visualization methods and tools that support selection of SaTScan parameters. Our methods distinguish between heterogeneous and homogeneous clusters and assess the stability of clusters across analytic scales. METHOD: We analyzed the cervical cancer mortality data for the United States aggregated by county between 2000 and 2004. We ran SaTScan on the dataset fifty times with different parameter choices. Our geovisual analytics approach couples SaTScan with our visual analytic platform, allowing users to interactively explore and compare SaTScan results produced by different parameter choices. The Standardized Mortality Ratio and reliability scores are visualized for all the counties to identify stable, homogeneous clusters. We evaluated our analysis result by comparing it to that produced by other independent techniques including the Empirical Bayes Smoothing and Kafadar spatial smoother methods. The geovisual analytics approach introduced here is developed and implemented in our Java-based Visual Inquiry Toolkit.