Data Visualization Saliency Model: A Tool for Evaluating Abstract Data Visualizations

Abstract

Evaluating the effectiveness of data visualizations is a challenging undertaking and often relies on one-off studies that test a visualization in the context of one specific task. Researchers across the fields of data science, visualization, and human-computer interaction are calling for foundational tools and principles that could be applied to assessing the effectiveness of data visualizations in a more rapid and generalizable manner. One possibility for such a tool is a model of visual saliency for data visualizations. Visual saliency models are typically based on the properties of the human visual cortex and predict which areas of a scene have visual features (e.g. color, luminance, edges) that are likely to draw a viewer's attention. While these models can accurately predict where viewers will look in a natural scene, they typically do not perform well for abstract data visualizations. In this paper, we discuss the reasons for the poor performance of existing saliency models when applied to data visualizations. We introduce the Data Visualization Saliency (DVS) model, a saliency model tailored to address some of these weaknesses, and we test the performance of the DVS model and existing saliency models by comparing the saliency maps produced by the models to eye tracking data obtained from human viewers. Finally, we describe how modified saliency models could be used as general tools for assessing the effectiveness of visualizations, including the strengths and weaknesses of this approach.

Visualization of 3D Property Data and Assessment of the Impact of Rendering Attributes

“…The systematic analysis of colour similarity and visual saliency emerged from some of the opinions on graphical styles stated in the initial interviews, but, within the scope of this study, no follow-up experimental study was conducted to compare the quantitative results with user's perceptions of different graphical styles and attributes. For the saliency analysis, such validation already exists partly based on gaze-tracking data (Matzen et al 2017), and for the colour similarity analysis, the presented trends in Fig. 5 agree well with our visual experience of corresponding renderings (Fig.…”

“…8f). The DVS model is, as the authors state, sensitive to low-level visual features that include not only colour stimuli but also local structural information such as present in line drawings and text (Matzen et al 2017). This results in the most transparent visualization in our example in comparably high levels of saliency across a large area of the visualization.…”

“…Many researchers agree that attention is not only stimulus-driven (Itti and Koch 2001;Parkhurst et al 2002;Underwood et al 2006). Still, the concept of visual saliency maps has been further improved recently, calibrated with eye-tracking studies in the field of data visualization (Matzen et al 2017), and has been suggested for the assessment of visualization designs (Wall et al 2019). In our study, we use the reference implementation of the Data Visualization Saliency (DVS) model by Wilson (2017) and described in Matzen et al (2017) to produce saliency maps of our visualizations under different conditions.…”

Quality Metrics for Information Visualization

“…For instance, Chen and Jänicke [JC10] proposed a method for computing a saliency‐based metric to measure the mismatches between visual salience and data characteristics (e.g., features detected by algorithms). Matzel at el [MHD*18] recently developed a saliency model to predict where people would look for a given visualization. Unlike models designed for images of natural scenes, their model attempts to incorporate top‐down visual features (e.g., texts) that are crucial for visualization tasks.…”

VIStory: interactive storyboard for exploring visual information in scientific publications

“…Specifically, in fields such as visualization, research processes generate imagery data that can substantially reflect content and quality of the research (Chen et al 2009). Taking visualization publications for an example, studying the visual information can benefit the field from multiple perspectives: to generate compact visual representations (e.g., Strobelt et al 2009), to guide design processes to make memorable, recognizable, and recallable visualizations (e.g., Borkin et al 2013Borkin et al , 2016, and to provide quality metrics for evaluating visualizations (e.g., Jänicke and Chen 2010;Matzen et al 2018).…”