Dashboards are one of the most common use cases for data visualization, and their design contexts considerably different from exploratory visualization tools. In this paper, we look at broad scope how dashboards used in practice through an analysis dashboard examples documentation about use. We systematically review literature surrounding use, construct a space dashboards, identify major types. characterize by goals, levels interaction, practices around them. Our framework suggest number...

Data surrounds each and every one of us in our daily lives, ranging from exercise logs, to archives interactions with others on social media, online resources pertaining hobbies. There is enormous potential for use these data understand ourselves better make positive changes lives. Visualization (Vis) visual analytics (VA) offer substantial opportunities help individuals gain insights about themselves, their communities interests; however, designing tools support analysis non-professional...

Natural language interfaces for visualizations have emerged as a promising new way of interacting with data and performing analytics. Many these systems fundamental limitations. Most return minimally interactive in response to queries often require experts perform modeling set predicted user before the are effective. Eviza provides natural interface an query dialog existing visualization rather than starting from blank sheet asking closed-ended questions that single text answer or static...

Designing collaborative interfaces for tabletops remains difficult because we do not fully understand how groups coordinate their actions when working collaboratively over tables. We present two observational studies of pairs completing independent and shared tasks that investigate coupling, or the manner in which collaborators are involved occupied with each other's work. Our results indicate individuals frequently fluidly engage disengage group activity through several distinct,...

It remains challenging for information visualization novices to rapidly construct visualizations during exploratory data analysis. We conducted an laboratory study in which explored fictitious sales by communicating specifications a human mediator, who constructed the using commercial software. found that three activities were central iterative construction process: attribute selection, visual template and mapping specification. The major barriers faced participants translating questions...

Visualization research generates beautiful images and impressive interactive systems. Emphasis on evaluating visualizations is growing. Researchers have successfully used alternative evaluation techniques in human-computer interaction (HCI), including focus groups, field studies, expert reviews. These methods tend to produce qualitative results require fewer participants than controlled experiments. In this article, we reviews that for the applications. We commonly use assess interface...

To verify cluster separation in high-dimensional data, analysts often reduce the data with a dimension reduction (DR) technique, and then visualize it 2D Scatterplots, interactive 3D or Scatterplot Matrices (SPLOMs). With goal of providing guidance between these visual encoding choices, we conducted an empirical study which two human coders manually inspected broad set 816 scatterplots derived from 75 datasets, 4 DR techniques, 3 previously mentioned scatterplot techniques. Each coder scored...

Abstract We provide two contributions, a taxonomy of visual cluster separation factors in scatterplots, and an in‐depth qualitative evaluation recently proposed validated measures. initially intended to use these measures guidance for the dimension reduction (DR) techniques encoding (VE) choices, but found that they failed produce reliable results. To understand why, we conducted systematic data study covering broad collection 75 real synthetic high‐dimensional datasets, four DR techniques,...

Interactive visual data analysis is most productive when users can focus on answering the questions they have about their data, rather than focusing how to operate interface tool. One viable approach engaging in interactive conversations with a natural language visualizations. These interfaces potential be both more expressive and accessible other interaction paradigms. We explore principles from pragmatics applied flow of analytical conversations, using as an input modality. evaluate...

Mobile devices have begun to raise questions around the potential for overuse when in presence of family or friends. As such, we conducted a diary and interview study understand how people use mobile others at home, this shapes their behavior household dynamics. Results show that members become frustrated do non-urgent activities on phones others. Yet often guess what are doing because personal nature devices. In some cases, developed strategies provide greater sense activity awareness...

When people work together to analyze a data set, they need organize their findings, hypotheses, and evidence, share that information with collaborators, coordinate activities amongst team members. Sharing externalizations (recorded such as notes) could increase awareness assist communication coordination. However, we currently know little about how provide tool support for this sort of sharing. We explore linked common (LCW) can be employed within `collaborative thinking space', facilitate...

We present the novel high-level visualization taxonomy. Our taxonomy classifies algorithms rather than data. Algorithms are categorized based on assumptions they make about data being visualized; we call this set of design model. Because our is models, it more flexible existing taxonomies and considers user's conceptual model, emphasizing human aspect visualization. Design models classified according to whether discrete or continuous by how much algorithm designer chooses display attributes...

We describe a series of experiments that compare 2D displays, 3D and combined 2D/3D displays (orientation icon, ExoVis, clip planes) for relative position estimation, orientation, volume interest tasks. Our results indicate can be very effective approximate navigation positioning when appropriate cues, such as shadows, are present. However, not precise except possibly in specific circumstances, instance, good viewing angles or measurement tools available. For tasks other situations,...

We present the novel high-level visualization taxonomy. Our taxonomy classifies algorithms rather than data. Algorithms are categorized based on assumptions they make about data being visualized; we call this set of design model. Because our is models, it more flexible existing taxonomies and considers user's conceptual model, emphasizing human aspect visualization. Design models classified according to whether discrete or continuous by how much algorithm designer chooses display attributes...

Visualization researchers and practitioners engaged in generating or evaluating designs are faced with the difficult problem of transforming questions asked actions taken by target users from domain-specific language context into more abstract forms. Existing task classifications aim to provide support for this endeavour providing a carefully delineated suite actions. Our experience is that bottom-up approach part challenge: low-level interpret without higher-level analysis goals process. To...

Handling ambiguity and underspecification of users' utterances is challenging, particularly for natural language interfaces that help with visual analytical tasks. Constraints in the underlying platform expectations high precision recall require thoughtful inferencing to generate useful responses. In this paper, we introduce a system resolve partial based on syntactic semantic constraints expressions. We extend best practices information visualization employ heuristics constrain solution...

When an organization chooses one course of action over alternatives, this task typically falls on a decision maker with relevant knowledge, experience, and understanding context. Decision makers rely data analysis, which is either delegated to analysts, or done their own. Often the combines data, likely uncertain incomplete, non-formalized knowledge within multi-objective problem space, weighing recommendations analysts broader contexts goals. As most past research in visual analytics has...

Today's AI systems for medical decision support often succeed on benchmark datasets in research papers but fail real-world deployment. This work focuses the making of sepsis, an acute life-threatening systematic infection that requires early diagnosis with high uncertainty from clinician. Our aim is to explore design requirements can clinical experts better decisions sepsis. The study begins a formative investigating why abandon existing AI-powered Sepsis predictive module their electrical...

DimStiller is a system for dimensionality reduction and analysis. It frames the task of understanding transforming input dimensions as series analysis steps where users transform data tables by chaining together different techniques, called operators, into pipelines expressions. The individual operators have controls views that are linked based on structure expression. Users interact with operator to tune parameter choices, immediate visual feedback guiding exploration local neighborhoods...

Data science is a rapidly growing discipline and organizations increasingly depend on data work. Yet the ambiguity around science, what it is, who scientists are can make difficult for visualization researchers to identify impactful research trajectories. We have conducted retrospective analysis of work workers as described within visualization, human computer interaction, literature. From this we synthesis comprehensive model that describes breakdown into nine distinct roles. summarise...

Chatbots have garnered interest as conversational interfaces for a variety of tasks. While general design guidelines exist chatbot interfaces, little work explores analytical chatbots that support conversing with data. We explore Gricean Maxims to help inform the basic effective interaction. also draw inspiration from natural language data exploration ambiguity and intent handling. ran Wizard Oz studies 30 participants evaluate user expectations text voice variants. Results identified...

Computing devices within current work and play environments are relatively static. As the number of 'networked' grows, as people their become more dynamic, situations will commonly arise where users wish to use 'that device there' instead navigating through traditional user interface widgets such lists. This paper describes a process for identifying pointing gesture using custom tags stylus called gesturePen. Implementation details this system provided along with qualitative quantitative...

We present a user interface, based on parallel coordinates, that facilitates exploration of volume data. By explicitly representing the visualization parameter space, interface provides an overview rendering options and enables users to easily explore different parameters. Rendered images are stored in integrated history bar backtracking previous options. Initial usability testing showed clear agreement between experts various backgrounds (usability, graphic design, visualization, medical...

We introduce eSeeTrack, an eye-tracking visualization prototype that facilitates exploration and comparison of sequential gaze orderings in a static or dynamic scene. It extends current data visualizations by extracting patterns orderings, displaying these way does not depend on the number fixations scene, enabling users to compare from two more sets eye-gaze data. Extracting such was very difficult with previous techniques. eSeeTrack combines timeline tree-structured visual representation...