VIZPLAN: A VISUAL ANALYTICS PLATFORM FOR THE ASSESSMENT OF MULTIDIMENSIONAL INDICATORS OVER TIME

ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022 7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia VIZPLAN: A VISUAL ANALYTICS PLATFORM FOR THE ASSESSMENT OF MULTIDIMENSIONAL INDICATORS OVER TIME L. Leplat1 , R. S. Torres1,3 ∗, D. Aspen1 , A. Amundsen2 1 NTNU – Norwegian University of Science and Technology, Ålesund, Norway - , (ricardo.torres, dina.aspen)@ntnu.no 2 United Future Lab Norway, Ålesund, Norway - 3 Wageningen University & Research, Wageningen, The Netherlands Commission IV, WG IV/9 KEY WORDS: Information visualization, visual analytics, dashboard, radial structure, temporal data, multidimensional data. ABSTRACT: In this paper, we introduce VizPlan, a new platform to support the assessment of multidimensional indicators over time. VizPlan includes a visualisation scheme based on a radial visual structure that allows the direct comparison of indicator values over time, a search tool to support the identification of entities whose indicators are similar to each other, and a clustering tool to group entities according to their indicator scores. VizPlan was designed and implemented to be flexible; it can be easily tailored to the visualization and analysis of any multidimensional temporal data. In this paper, the use of VizPlan is illustrated in the context of three case studies concerning the analysis of sustainability indicators to support urban planning: key performance indicators related to the sustainable development goals, walkability analysis, and bus service availability assessment. All case studies refer to real data related to Norwegian cities, especially Ålesund. VizPlan is available as an open source software at https://github.com/Rylern/VizPlan – As of May 2022. 1. INTRODUCTION Technological advances have fostered the wide adoption of procedures for data acquisition, storage, and communication. A plethora of Smart City applications now relies on the analysis of huge volumes of data (Psyllidis et al., 2015, Costa and Santos, 2017, Mehmood et al., 2019, Zheng et al., 2016). In particular, special interest has been given to the analysis of different “whatif” scenarios based on the use of indicators computed from results obtained from sensors, surveys, and even simulation or data-driven methods. The proper assessment of time-varying indicators is fundamental for decision-making and, therefore, the existence of suitable visual structures to support the understanding of patterns across space and time (including changes) plays a relevant role (Zheng et al., 2016). Our study aims to understand how to support the analysis of multidimensional numerical indicators with temporal variability. In this paper, we introduce VizPlan, a visual analytics platform for the assessment of multidimensional indicators over time. VizPlan includes a radial visual structure that allows the direct comparison of indicator scores for different categories, for example, in the analysis of smart city key performance indicators (KPI) based on sustainable development goals. The developed platform supports map-based navigation, selection of indicators, and their comparison across different years. In a typical usage scenario, cities that have available KPI data can be displayed on an interactive map. Users can then select one of the cities by clicking on a map location to visualize its respective KPI chart. Other features included in the interactive chart refer to filtering on performance values, zooming, panning, label toggling, and the ability to compare segments against other cities. The platform also supports clustering and similarity searches, which allow the identification of entities (e.g., regions, cities) that are similar to each other according to a pre-defined set of indicators. The platform was designed and implemented to make it flexible. That platform can be easily tailored to different applications and problems. We have demonstrated its use in the context of three appealing applications related to decision-making based on the analysis of indicator changes over time: assessment of KPI indicators related to Sustainable Dvelopment Goals (SDG), evaluation of walkability properties in different regions of a city, and the analysis of bus service availability in several neighborhoods. The considered applications are linked to SDG 11 (target 11.2), which refers to providing “safe, affordable, accessible, and sustainable transport systems.”1 This paper is organized as follows: Section 2 briefly covers relevant related work; Section 3 introduces VizPlan, highlighting its main components and features; Section 4 illustrates the use of VizPlan in three case studies; finally, Section 5 outlines our conclusions and presents directions for future studies. 2. RELATED WORK The literature is vast with regard to studies related to the visualization of multidimensional data (Ltifi et al., 2020, Liu et al., 2017). Popular strategies rely on the use of scatter plots (Friendly and Denis, 2005), radar charts (Albo et al., 2016), and parallel coordinates (Inselberg, 1985). Another successful strategy relies on the use of radial structures (Draper et al., 2009, Albo et al., 2016). Encoding and representing changes of multidimensional data over time has also been investigated in several applications, especially for urban data (Zheng et al., 2016). For more details regarding successful approaches for time-oriented 1 ∗ Corresponding author https://unstats.un.org/sdgs/metadata/?Text=&Goal= &Target=11.2 (As of July 2022). This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-127-2022 | © Author(s) 2022. CC BY 4.0 License. 127 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022 7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia data visualization, the readers may refer to (Aigner et al., 2008, Aigner et al., 2011). Mariano et al. (Mariano et al., 2018, Mariano et al., 2019), for example, integrated visual rhythms with radial structures to support the analysis of phenological data encoded in stack of relational tables (Mariano et al., 2018) or images (Mariano et al., 2019). The main goal was to support the analysis of temporal changes of phenological variables. In the current version of VizPlan, we utilize a similar strategy for the visualization of indicators associated with different geographical objects. In the context of urban data analysis, information visualization approaches have been employed to several applications, such as the assessment of mobility data (Feng et al., 2021), water source management (Xu et al., 2022), urban pollution (Bello et al., 2019), and land use evolution (Santos et al., 2021). In broader formulations, data integration infrastructur (...truncated)