Visualizations of tabular data are widely used; understanding their effectiveness in different task and contexts is fundamental to scaling impact. However, little known about how basic visualizations perform across varying analysis tasks. In this paper, we report results from a crowdsourced experiment evaluate the five small scale (5-34 points) two-dimensional visualization types-Table, Line Chart, Bar Scatterplot, Pie Chart-across ten common tasks using two datasets. We find these types...

Traditionally, studies of data visualization techniques and systems have evaluated visualizations with respect to usability goals such as effectiveness efficiency. These assess performance-related metrics time correctness participants completing analytic tasks. Alternatively, several in InfoVis recently by investigating user experience memorability, engagement, enjoyment fun. employ somewhat different evaluation methodologies these other goals. The growing number studies, their alternative...

Although data visualization tools continue to improve, during the exploration process many of them require users manually specify techniques, mappings, and parameters. In response, we present Visualization by Demonstration paradigm, a novel interaction method for visual exploration. A system which adopts this paradigm allows provide demonstrations incremental changes representation. The then recommends potential transformations (Visual Representation, Data Mapping, Axes, View Specification...

We conducted an experiment to understand how mobile phone users perceive the urgency of ten simple vibration alerts that were created from four basic signals: short on, off, long and off. The signals correspond 200 ms 600 ms, respectively. To convey level notifications help prioritize them, design should consider gap length preceding or succeeding a signal, number gaps in pattern, vibration's duration affect alert's perceived urgency. Our study specifically shows shorter lengths between...

Abstract We investigate the memorability of data represented in two different visualization designs. In contrast to recent studies that examine which types visual information make visualizations memorable, we effect on time and accuracy recall displayed , minutes days after interaction with visualizations. particular, describe results an evaluation comparing same relational data: node‐link diagrams map‐based visualization. find significant differences tasks performed, these persist original...

Effectively showing the relationships between objects in a dataset is one of main tasks information visualization. Typically there well-defined notion distance pairs objects, and traditional approaches such as principal component analysis or multi-dimensional scaling are used to place points 2D space, so that similar close each other. In another typical setting, visualized network graph, where related nodes connected by links. More recently, datasets also maps, addition links, an explicit...

User interfaces for data visualization often consist of two main components: control panels user interaction and visual representation. A recent trend in is directly embedding into the representations. For example, instead using to adjust parameters, users can basic graphical encodings (e.g., changing distances between points a scatterplot) perform similar parameterizations. However, enabling embedded interactions requires strong understanding how influence ability accurately perceive...

Abstract While evaluation studies in visualization often involve traditional performance measurements, there has been a concerted effort to move beyond time and accuracy. Of these alternative aspects, memorability recall of visualizations have recently considered, but other aspects such as enjoyment engagement are not well explored. We study the two different methods through user study. In particular, we describe results three‐phase experiment comparing same relational data: node‐link...

Many visual analytics systems allow users to interact with machine learning models towards the goals of data exploration and insight generation on a given dataset. However, in some situations, insights may be less important than production an accurate predictive model for future use. In that case, are more interested generating diverse robust models, verifying their performance holdout data, selecting most suitable usage scenario. this paper, we consider concept Exploratory Model Analysis...

Enterprise networks have been a frequent target of data breaches and sabotage. In widely used method, attackers establish foothold in the network by compromising single computer or account. They then move laterally between computers to access valuable resources information located deeper inside network. To laterally, often steal valid user credentials. This paper is based on observation that an attackers' pattern characteristics stolen credentials form <;User, Source, Destination> deviates...

Task taxonomies for graph and network visualizations focus on tasks commonly encountered when analyzing connectivity topology. However, in many application fields such as the social sciences (social networks), biology (protein interaction models), software engineering (program call graphs), topology information is intertwined with group, clustering, hierarchical information. Several recent visualization techniques, BubbleSets, LineSets GMap, make explicit use of grouping but evaluating has...

In this paper, we describe a research agenda for deriving design principles directly from data. We argue that it is time to go beyond manually curated and applied visualization guidelines. propose learning models of data collected using graphical perception studies build tools powered by the learned models. To achieve vision, need 1) develop scalable methods collecting training data, 2) collect different forms 3) advance interpretability machine models, 4) adaptive evolve as more becomes available.

Task taxonomies for graph and network visualization focus on tasks commonly encountered when analyzing connectivity topology. However, in many application fields such as the social sciences (social networks), biology (protein interaction models), software engineering (program call graphs), topology information is intertwined with grouping clustering information. Several recent techniques, BubbleSets, LineSets GMap, make explicit use of clustering, but evaluating visualizations has been...

We investigate direct manipulation of graphical encodings as a method for interacting with visualizations. There is an increasing interest in developing visualization tools that enable users to perform operations by directly manipulating rather than external widgets such checkboxes and sliders. Designers must decide which should be supported, identify how each operation can invoked. However, we lack empirical guidelines people convey their intended using encodings. address this issue...

Previous studies have suggested that social media data, along with machine learning algorithms, can be used to generate computational mental health insights. These insights the potential support clinician-patient communication during psychotherapy consultations. However, how clinicians perceive and envision using consultations has been underexplored.The aim of this study is understand clinician perspectives regarding from patients' activities. We focus on opportunities challenges these...

Traditionally, evaluation studies in information visualization have measured effectiveness by assessing performance time and accuracy. More recently, there has been a concerted effort to understand aspects beyond errors. In this paper we study enjoyment, which, while arguably not the primary goal of visualization, shown impact memorability. Different models enjoyment proposed psychology, education gaming; yet is no standard approach evaluate measure visualization. relate flow model...

Recently, there has been an increasing trend to extend the demonstrational interaction paradigm visualization tools. As more analytic operations can be performed by demonstration, new user tasks supported. In this paper, we discuss properties of where by-demonstration effective and describe main components needed implement in

Biologists often perform clustering analysis to derive meaningful patterns, relationships, and structures from data instances attributes. Though plays a pivotal role in biologists' exploration, it takes non-trivial efforts for biologists find the best grouping their using existing tools. Visual cluster is currently performed either programmatically or through menus dialogues many tools, which require parameter adjustments over several steps of trial-and-error. In this article, we introduce...

Abstract A comprehensive understanding of collocated social interactions can help campuses and organizations better support their community. Universities could determine new ways to conduct classes design programs by studying how students have in the past. However, this needs data that describe large groups over a long period. Harnessing user devices infer collocation, while tempting, is challenged privacy concerns, power consumption, maintenance issues. Alternatively, embedding sensors...

Static visualizations have analytic and expressive value. However, many interactive tasks cannot be completed using static visualizations. As datasets grow in size complexity, start losing their power for data exploration. Despite this limitation of visualizations, there are still cases where limited to being (e.g., on presentation slides or posters). We believe these cases, will benefit from allowing users perform them. Inspired by the introduction numerous commercial personal augmented...

Managing time while presenting is challenging, but mobile devices offer both convenience and flexibility in their ability to support the end-to-end process of setting, refining, following presentation targets. From an initial HCI-Q study 20 presenters, we identified need set such targets per 'zone' consecutive slides (rather than slide or for whole talk), as well feedback that accommodates two distinct attitudes management. These findings led design TalkZones, a application timing support....

Visualizations of tabular data are widely used; understanding their effectiveness in different task and contexts is fundamental to scaling impact. However, little known about how basic visualizations perform across varying analysis tasks attribute types. In this paper, we report results from a crowdsourced experiment evaluate the five visualization types --- Table, Line Chart, Bar Scatterplot, Pie Chart ten common three using two real-world datasets. We found these significantly varies...