IEEE VIS Publication Dataset

RadViz and star coordinates are two of the most popular projection-based multivariate visualization techniques that arrange variables in radial layouts. Formally, the main difference between them consists of a nonlinear normalization step inherent in RadViz. In this paper we show that, although RadViz can be useful when analyzing sparse data, in general this design choice limits its applicability and introduces several drawbacks for exploratory data analysis. In particular, we observe that the normalization step introduces nonlinear distortions, can encumber outlier detection, prevents associating the plots with useful linear mappings, and impedes estimating original data attributes accurately. In addition, users have greater flexibility when choosing different layouts and views of the data in star coordinates. Therefore, we suggest that analysts and researchers should carefully consider whether RadViz's normalization step is beneficial regarding the data sets' characteristics and analysis tasks.

When data categories have strong color associations, it is useful to use these semantically meaningful concept-color associations in data visualizations. In this paper, we explore how linguistic information about the terms defining the data can be used to generate semantically meaningful colors. To do this effectively, we need first to establish that a term has a strong semantic color association, then discover which color or colors express it. Using co-occurrence measures of color name frequencies from Google n-grams, we define a measure for colorability that describes how strongly associated a given term is to any of a set of basic color terms. We then show how this colorability score can be used with additional semantic analysis to rank and retrieve a representative color from Google Images. Alternatively, we use symbolic relationships defined by WordNet to select identity colors for categories such as countries or brands. To create visually distinct color palettes, we use k-means clustering to create visually distinct sets, iteratively reassigning terms with multiple basic color associations as needed. This can be additionally constrained to use colors only in a predefined palette.

Physical visualizations, or data physicalizations, encode data in attributes of physical shapes. Despite a considerable body of work on visual variables, ΓÇ£physical variablesΓÇ¥ remain poorly understood. One of them is physical size. A difficulty for solid elements is that ΓÇ£sizeΓÇ¥ is ambiguous - it can refer to either length/diameter, surface, or volume. Thus, it is unclear for designers of physicalizations how to effectively encode quantities in physical size. To investigate, we ran an experiment where participants estimated ratios between quantities represented by solid bars and spheres. Our results suggest that solid bars are compared based on their length, consistent with previous findings for 2D and 3D bars on flat media. But for spheres, participants' estimates are rather proportional to their surface. Depending on the estimation method used, judgments are rather consistent across participants, thus the use of perceptually-optimized size scales seems possible. We conclude by discussing implications for the design of data physicalizations and the need for more empirical studies on physical variables.

General methods for drawing Euler diagrams tend to generate irregular polygons. Yet, empirical evidence indicates that smoother contours make these diagrams easier to read. In this paper, we present a simple method to smooth the boundaries of any Euler diagram drawing. When refining the diagram, the method must ensure that set elements remain inside their appropriate boundaries and that no region is removed or created in the diagram. Our approach uses a force system that improves the diagram while at the same time ensuring its topological structure does not change. We demonstrate the effectiveness of the approach through case studies and quantitative evaluations.

While information visualization frameworks and heuristics have traditionally been reluctant to include acquired codes of meaning, designers are making use of them in a wide variety of ways. Acquired codes leverage a user's experience to understand the meaning of a visualization. They range from figurative visualizations which rely on the reader's recognition of shapes, to conventional arrangements of graphic elements which represent particular subjects. In this study, we used content analysis to codify acquired meaning in visualization. We applied the content analysis to a set of infographics and data visualizations which are exemplars of innovative and effective design. 88% of the infographics and 71% of data visualizations in the sample contain at least one use of figurative visualization. Conventions on the arrangement of graphics are also widespread in the sample. In particular, a comparison of representations of time and other quantitative data showed that conventions can be specific to a subject. These results suggest that there is a need for information visualization research to expand its scope beyond perceptual channels, to include social and culturally constructed meaning. Our paper demonstrates a viable method for identifying figurative techniques and graphic conventions and integrating them into heuristics for visualization design.

Datasets commonly include multi-value (set-typed) attributes that describe set memberships over elements, such as genres per movie or courses taken per student. Set-typed attributes describe rich relations across elements, sets, and the set intersections. Increasing the number of sets results in a combinatorial growth of relations and creates scalability challenges. Exploratory tasks (e.g. selection, comparison) have commonly been designed in separation for set-typed attributes, which reduces interface consistency. To improve on scalability and to support rich, contextual exploration of set-typed data, we present AggreSet. AggreSet creates aggregations for each data dimension: sets, set-degrees, set-pair intersections, and other attributes. It visualizes the element count per aggregate using a matrix plot for set-pair intersections, and histograms for set lists, set-degrees and other attributes. Its non-overlapping visual design is scalable to numerous and large sets. AggreSet supports selection, filtering, and comparison as core exploratory tasks. It allows analysis of set relations inluding subsets, disjoint sets and set intersection strength, and also features perceptual set ordering for detecting patterns in set matrices. Its interaction is designed for rich and rapid data exploration. We demonstrate results on a wide range of datasets from different domains with varying characteristics, and report on expert reviews and a case study using student enrollment and degree data with assistant deans at a major public university.

Node-link diagrams provide an intuitive way to explore networks and have inspired a large number of automated graph layout strategies that optimize aesthetic criteria. However, any particular drawing approach cannot fully satisfy all these criteria simultaneously, producing drawings with visual ambiguities that can impede the understanding of network structure. To bring attention to these potentially problematic areas present in the drawing, this paper presents a technique that highlights common types of visual ambiguities: ambiguous spatial relationships between nodes and edges, visual overlap between community structures, and ambiguity in edge bundling and metanodes. Metrics, including newly proposed metrics for abnormal edge lengths, visual overlap in community structures and node/edge aggregation, are proposed to quantify areas of ambiguity in the drawing. These metrics and others are then displayed using a heatmap-based visualization that provides visual feedback to developers of graph drawing and visualization approaches, allowing them to quickly identify misleading areas. The novel metrics and the heatmap-based visualization allow a user to explore ambiguities in graph layouts from multiple perspectives in order to make reasonable graph layout choices. The effectiveness of the technique is demonstrated through case studies and expert reviews.

Effective small multiple displays are created by partitioning a visualization on variables that reveal interesting conditional structure in the data. We propose a method that automatically ranks partitioning variables, allowing analysts to focus on the most promising small multiple displays. Our approach is based on a randomized, non-parametric permutation test, which allows us to handle a wide range of quality measures for visual patterns defined on many different visualization types, while discounting spurious patterns. We demonstrate the effectiveness of our approach on scatterplots of real-world, multidimensional datasets.

In this paper we move beyond memorability and investigate how visualizations are recognized and recalled. For this study we labeled a dataset of 393 visualizations and analyzed the eye movements of 33 participants as well as thousands of participant-generated text descriptions of the visualizations. This allowed us to determine what components of a visualization attract people's attention, and what information is encoded into memory. Our findings quantitatively support many conventional qualitative design guidelines, including that (1) titles and supporting text should convey the message of a visualization, (2) if used appropriately, pictograms do not interfere with understanding and can improve recognition, and (3) redundancy helps effectively communicate the message. Importantly, we show that visualizations memorable ΓÇ£at-a-glanceΓÇ¥ are also capable of effectively conveying the message of the visualization. Thus, a memorable visualization is often also an effective one.

Models of human perception - including perceptual ΓÇ£lawsΓÇ¥ - can be valuable tools for deriving visualization design recommendations. However, it is important to assess the explanatory power of such models when using them to inform design. We present a secondary analysis of data previously used to rank the effectiveness of bivariate visualizations for assessing correlation (measured with Pearson's r) according to the well-known Weber-Fechner Law. Beginning with the model of Harrison et al. [1], we present a sequence of refinements including incorporation of individual differences, log transformation, censored regression, and adoption of Bayesian statistics. Our model incorporates all observations dropped from the original analysis, including data near ceilings caused by the data collection process and entire visualizations dropped due to large numbers of observations worse than chance. This model deviates from Weber's Law, but provides improved predictive accuracy and generalization. Using Bayesian credibility intervals, we derive a partial ranking that groups visualizations with similar performance, and we give precise estimates of the difference in performance between these groups. We find that compared to other visualizations, scatterplots are unique in combining low variance between individuals and high precision on both positively- and negatively correlated data. We conclude with a discussion of the value of data sharing and replication, and share implications for modeling similar experimental data.

The parallel coordinates technique is widely used for the analysis of multivariate data. During recent decades significant research efforts have been devoted to exploring the applicability of the technique and to expand upon it, resulting in a variety of extensions. Of these many research activities, a surprisingly small number concerns user-centred evaluations investigating actual use and usability issues for different tasks, data and domains. The result is a clear lack of convincing evidence to support and guide uptake by users as well as future research directions. To address these issues this paper contributes a thorough literature survey of what has been done in the area of user-centred evaluation of parallel coordinates. These evaluations are divided into four categories based on characterization of use, derived from the survey. Based on the data from the survey and the categorization combined with the authors' experience of working with parallel coordinates, a set of guidelines for future research directions is proposed.

Semi-automatic text analysis involves manual inspection of text. Often, different text annotations (like part-of-speech or named entities) are indicated by using distinctive text highlighting techniques. In typesetting there exist well-known formatting conventions, such as bold typeface, italics, or background coloring, that are useful for highlighting certain parts of a given text. Also, many advanced techniques for visualization and highlighting of text exist; yet, standard typesetting is common, and the effects of standard typesetting on the perception of text are not fully understood. As such, we surveyed and tested the effectiveness of common text highlighting techniques, both individually and in combination, to discover how to maximize pop-out effects while minimizing visual interference between techniques. To validate our findings, we conducted a series of crowd-sourced experiments to determine: i) a ranking of nine commonly-used text highlighting techniques; ii) the degree of visual interference between pairs of text highlighting techniques; iii) the effectiveness of techniques for visual conjunctive search. Our results show that increasing font size works best as a single highlighting technique, and that there are significant visual interferences between some pairs of highlighting techniques. We discuss the pros and cons of different combinations as a design guideline to choose text highlighting techniques for text viewers.

Prior research into network layout has focused on fast heuristic techniques for layout of large networks, or complex multi-stage pipelines for higher quality layout of small graphs. Improvements to these pipeline techniques, especially for orthogonal-style layout, are difficult and practical results have been slight in recent years. Yet, as discussed in this paper, there remain significant issues in the quality of the layouts produced by these techniques, even for quite small networks. This is especially true when layout with additional grouping constraints is required. The first contribution of this paper is to investigate an ultra-compact, grid-like network layout aesthetic that is motivated by the grid arrangements that are used almost universally by designers in typographical layout. Since the time when these heuristic and pipeline-based graph-layout methods were conceived, generic technologies (MIP, CP and SAT) for solving combinatorial and mixed-integer optimization problems have improved massively. The second contribution of this paper is to reassess whether these techniques can be used for high-quality layout of small graphs. While they are fast enough for graphs of up to 50 nodes we found these methods do not scale up. Our third contribution is a large-neighborhood search meta-heuristic approach that is scalable to larger networks.

Over the last 50 years a wide variety of automatic network layout algorithms have been developed. Some are fast heuristic techniques suitable for networks with hundreds of thousands of nodes while others are multi-stage frameworks for higher-quality layout of smaller networks. However, despite decades of research currently no algorithm produces layout of comparable quality to that of a human. We give a new ΓÇ£human-centredΓÇ¥ methodology for automatic network layout algorithm design that is intended to overcome this deficiency. User studies are first used to identify the aesthetic criteria algorithms should encode, then an algorithm is developed that is informed by these criteria and finally, a follow-up study evaluates the algorithm output. We have used this new methodology to develop an automatic orthogonal network layout method, HOLA, that achieves measurably better (by user study) layout than the best available orthogonal layout algorithm and which produces layouts of comparable quality to those produced by hand.

In this paper, we would like to investigate how people make sense of unfamiliar information visualizations. In order to achieve the research goal, we conducted a qualitative study by observing 13 participants when they endeavored to make sense of three unfamiliar visualizations (i.e., a parallel-coordinates plot, a chord diagram, and a treemap) that they encountered for the first time. We collected data including audio/video record of think-aloud sessions and semi-structured interview; and analyzed the data using the grounded theory method. The primary result of this study is a grounded model of NOvice's information VIsualization Sensemaking (NOVIS model), which consists of the five major cognitive activities: 1 encountering visualization, 2 constructing a frame, 3 exploring visualization, 4 questioning the frame, and 5 floundering on visualization. We introduce the NOVIS model by explaining the five activities with representative quotes from our participants. We also explore the dynamics in the model. Lastly, we compare with other existing models and share further research directions that arose from our observations.

Decades of research have repeatedly shown that people perform poorly at estimating and understanding conditional probabilities that are inherent in Bayesian reasoning problems. Yet in the medical domain, both physicians and patients make daily, life-critical judgments based on conditional probability. Although there have been a number of attempts to develop more effective ways to facilitate Bayesian reasoning, reports of these findings tend to be inconsistent and sometimes even contradictory. For instance, the reported accuracies for individuals being able to correctly estimate conditional probability range from 6% to 62%. In this work, we show that problem representation can significantly affect accuracies. By controlling the amount of information presented to the user, we demonstrate how text and visualization designs can increase overall accuracies to as high as 77%. Additionally, we found that for users with high spatial ability, our designs can further improve their accuracies to as high as 100%. By and large, our findings provide explanations for the inconsistent reports on accuracy in Bayesian reasoning tasks and show a significant improvement over existing methods. We believe that these findings can have immediate impact on risk communication in health-related fields.

The energy performance of large building portfolios is challenging to analyze and monitor, as current analysis tools are not scalable or they present derived and aggregated data at too coarse of a level. We conducted a visualization design study, beginning with a thorough work domain analysis and a characterization of data and task abstractions. We describe generalizable visual encoding design choices for time-oriented data framed in terms of matches and mismatches, as well as considerations for workflow design. Our designs address several research questions pertaining to scalability, view coordination, and the inappropriateness of line charts for derived and aggregated data due to a combination of data semantics and domain convention. We also present guidelines relating to familiarity and trust, as well as methodological considerations for visualization design studies. Our designs were adopted by our collaborators and incorporated into the design of an energy analysis software application that will be deployed to tens of thousands of energy workers in their client base.

A composite indicator (CI) is a measuring and benchmark tool used to capture multi-dimensional concepts, such as Information and Communication Technology (ICT) usage. Individual indicators are selected and combined to reflect a phenomena being measured. Visualization of a composite indicator is recommended as a tool to enable interested stakeholders, as well as the public audience, to better understand the indicator components and evolution overtime. However, existing CI visualizations introduce a variety of solutions and there is a lack in CI's visualization guidelines. Radial visualizations are popular among these solutions because of CI's inherent multi-dimensionality. Although in dispute, Radar-charts are often used for CI presentation. However, no empirical evidence on Radar's effectiveness and efficiency for common CI tasks is available. In this paper, we aim to fill this gap by reporting on a controlled experiment that compares the Radar chart technique with two other radial visualization methods: Flowercharts as used in the well-known OECD Betterlife index, and Circle-charts which could be adopted for this purpose. Examples of these charts in the current context are shown in Figure 1. We evaluated these charts, showing the same data with each of the mentioned techniques applying small multiple views for different dimensions of the data. We compared users' performance and preference empirically under a formal task-taxonomy. Results indicate that the Radar chart was the least effective and least liked, while performance of the two other options were mixed and dependent on the task. Results also showed strong preference of participants toward the Flower chart. Summarizing our results, we provide specific design guidelines for composite indicator visualization.

Finding good projections of n-dimensional datasets into a 2D visualization domain is one of the most important problems in Information Visualization. Users are interested in getting maximal insight into the data by exploring a minimal number of projections. However, if the number is too small or improper projections are used, then important data patterns might be overlooked. We propose a data-driven approach to find minimal sets of projections that uniquely show certain data patterns. For this we introduce a dissimilarity measure of data projections that discards affine transformations of projections and prevents repetitions of the same data patterns. Based on this, we provide complete data tours of at most n/2 projections. Furthermore, we propose optimal paths of projection matrices for an interactive data exploration. We illustrate our technique with a set of state-of-the-art real high-dimensional benchmark datasets.

Parallel Coordinate Plots (PCPs) is one of the most powerful techniques for the visualization of multivariate data. However, for large datasets, the representation suffers from clutter due to overplotting. In this case, discerning the underlying data information and selecting specific interesting patterns can become difficult. We propose a new and simple technique to improve the display of PCPs by emphasizing the underlying data structure. Our Orientation-enhanced Parallel Coordinate Plots (OPCPs) improve pattern and outlier discernibility by visually enhancing parts of each PCP polyline with respect to its slope. This enhancement also allows us to introduce a novel and efficient selection method, the Orientation-enhanced Brushing (O-Brushing). Our solution is particularly useful when multiple patterns are present or when the view on certain patterns is obstructed by noise. We present the results of our approach with several synthetic and real-world datasets. Finally, we conducted a user evaluation, which verifies the advantages of the OPCPs in terms of discernibility of information in complex data. It also confirms that O-Brushing eases the selection of data patterns in PCPs and reduces the amount of necessary user interactions compared to state-of-the-art brushing techniques.