The information visualization research community commonly produces supporting software to demonstrate technical contributions to the field. However, developing this software tends to be an overwhelming task. The final product tends to be a research prototype without much thought for modularization and re-usability, which makes it harder to replicate and adopt. This paper presents a design pattern for facilitating the creation, dissemination, and re-utilization of visualization techniques using reactive widgets. The design pattern features basic concepts that leverage modern front-end development best practices and standards, which facilitate development and replication. The paper presents several usage examples of the pattern, templates for implementation, and even a wrapper for facilitating the conversion of any Vega [27], [28] specification into a reactive widget. https://johnguerra.co/reactiveWidgets

Behaviour in virtual environments might be informed by our experiences in physical environments, but virtual environments are not constrained by the same physical, perceptual, or social cues. Instead of replicating the properties of physical spaces, one can create virtual experiences that diverge from reality by dynamically manipulating environmental, aural, and social properties. This paper explores digital proxemics, which describe how we use space in virtual environments and how the presence of others influences our behaviours, interactions, and movements. First, we frame the open challenges of digital proxemics in terms of activity, social signals, audio design, and environment. We explore a subset of these challenges through an evaluation that compares two audio designs and two displays with different social signal affordances: head-mounted display (HMD) versus desktop PC. We use quantitative methods using instrumented tracking to analyse behaviour, demonstrating how personal space, proximity, and attention compare between desktop PC and HMDs.

In the field of information visualization, the concept of "tasks" is an essential component of theories and methodologies for how a visualization researcher or a practitioner understands what tasks a user needs to perform and how to approach the creation of a new design. In this article, we focus on the collection of tasks for tree visualizations, a common visual encoding in many domains ranging from biology to computer science to geography. In spite of their commonality, no prior efforts exist to collect and abstractly define tree visualization tasks. We present a literature review of tree visualization articles and generate a curated dataset of over 200 tasks. To enable effective task abstraction for trees, we also contribute a novel extension of the Multi-Level Task Typology to include more specificity to support tree-specific tasks as well as a systematic procedure to conduct task abstractions for tree visualizations. All tasks in the dataset were abstracted with the novel typology extension and analyzed to gain a better understanding of the state of tree visualizations. These abstracted tasks can benefit visualization researchers and practitioners as they design evaluation studies or compare their analytical tasks with ones previously studied in the literature to make informed decisions about their design. We also reflect on our novel methodology and advocate more broadly for the creation of task-based knowledge repositories for different types of visualizations. The Supplemental Material, which can be found on the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TVCG.2021.3064037, will be maintained on OSF: https://osf.io/u5ehs/.

Tactile graphics (TG) can help people with visual disabilities access visual concepts. However, the number of TGs available to users is considerably limited because they need to be created by designers and teachers of the visually impaired (TVIs) with extensive experience. High-quality images can be transformed into TGs. In order to increase the availability of TGs, we trained a machine learning (ML) model that identifies suitable and unsuitable images for TG transformation (See Figure 1). This model would help users identify high-quality images that can be transformed into TGs. The poster presents (1) the ML model trained with 800 images collected from the American Printing House tactile Library and the researchers, (2) a web application that lets TVIs retrain the model by feeding new images and helping with the classification. This system can then be used by anyone, especially parents and teachers, as a filter to produce new TGs.

The CWI-ADE2016 Dataset is a collection of more than 40 million Bluetooth Low Energy (BLE) packets and of 14 million accelerometer and temperature samples generated by wristbands that people wore in a nightclub. The data was gathered during Amsterdam Dance Event 2016 in an exclusive club experience curated around human senses, which leveraged technology as a bridge between the club and the guests. Each guest was handed a custom-made wristband with a BLE-enabled device that broadcast movement, temperature and other sensor readings. A network of Raspberry Pi receivers deployed for the occasion captured broadcast packets from wristbands and any other BLE device in the environment. This data provides a full picture of the performance of the real life deployment of a sensing infrastructure and gives insights to designing sensing platforms, understanding networks and crowds behaviour or studying opportunistic sensing. This paper describes an analysis of this dataset and some examples of usage.

Online photo collections are often presented along with numeric data, such as views, likes, or comments. Treemaps are ideal for visualizing such collections, as they present numeric values using rectangular blocks, within which photos can be presented. Despite abundant research showing that photo treemaps and similar space-filling approaches are useful and appealing for users, understanding of the ideal parameters for constructing these representations remains limited. To address this, we contribute a series of experiments targeted at identifying design parameters for building photo treemaps. Our first study explores the number of photos presented at each level in the treemap, finding that using fewer photos makes visualizations more searchable and preferable to users. A second study examines the tradeoff between sizing photos according to the numeric values and presenting them in a more familiar layout with fixed aspect ratios. These results inform a third study, in which a prototype was used to probe the preferences of active Flickr users for using treemaps for navigating photo collections.

This paper describes the process of design, implementation, and real world deployment of a web-based network exploration tool called Network Explorer. We designed Network Explorer based on the expressed needs of our clients and later deployed it as part of a larger system for fraud detection in health care. Our implementation of Network Explorer provides visual interactive access to large-scale network data. As part of the Network Explorer tool we contribute a dynamic group-in-a-box implementation for laying out clusters, and a node navigator widget that aids in the exploration of large networks. We are also contributing two open source components of the Network Explorer for the community to reuse: an in-browser clustering library, and the dynamic group-in-a-box algorithm. We have evaluated the network explorer tool in multiple real-world environments including the fraud detection setting above.

Temporal terms, such as 'winter', 'Christmas', or 'January' are often used in search queries for personal images. But how do people's memories and perceptions of time match with the actual dates when their images were captured? We compared the temporal terms that 74 Flickr users used to search their own photo collections, and compared them to the date captured data in the target image. We also conducted a larger study across several billion images, comparing user-applied tags for holidays and seasons to the dates the images were captured. We demonstrate that various query terms and tags can be in conflict with the actual dates photos were taken for specific types of temporal terms up to 40% of the time. We will conclude by highlighting implications for search systems where users are querying for personal content by date.

Participants in social media systems must balance many considerations when choosing what to share and with whom. Sharing with others invites certain risks, as well as potential benefits; achieving the right balance is even more critical when sharing photos, which can be particularly engaging, but potentially compromising. In this paper, we examine photo-sharing decisions as an interaction between high-level user preferences and specific features of the images being shared. Our analysis combines insights from a 96-user survey with metadata from 10.4M photos to develop a model integrating these perspectives to predict permissions settings for uploaded photos. We discuss implications, including how such a model can be applied to provide online sharing experiences that are more safe, more scalable, and more satisfying.

Detection of fraud, waste, and abuse (FWA) is an important yet difficult problem. In this paper, we describe a system to detect suspicious activities in large healthcare claims datasets. Each healthcare dataset is viewed as a heterogeneous network of patients, doctors, pharmacies, and other entities. These networks can be large, with millions of patients, hundreds of thousands of doctors, and tens of thousands of pharmacies, for example. Graph analysis techniques are developed to find suspicious individuals, suspicious relationships between individuals, unusual changes over time, unusual geospatial dispersion, and anomalous networks within the overall graph structure. The system has been deployed on multiple sites and data sets, both government and commercial, to facilitate the work of FWA investigation analysts.

To analyze data such as the US Federal Budget or characteristics of the student population of a University it is common to look for changes over time. This task can be made easier and more fruitful if the analysis is performed by grouping by attributes, such as by Agencies, Bureaus and Accounts for the Budget, or Ethnicity, Gender and Major in a University. We present TreeVersity2, a web based interactive data visualization tool that allows users to analyze change in datasets by creating dynamic hierarchies based on the data attributes. TreeVersity2 introduces a novel space filling visualization (StemView) to represent change in trees at multiple levels--not just at the leaf level. With this visualization users can explore absolute and relative changes, created and removed nodes, and each node's actual values, while maintaining the context of the tree. In addition, TreeVersity2 provides overviews of change over the entire time period, and a reporting tool that lists outliers in textual form, which helps users identify the major changes in the data without having to manually setup filters. We validated TreeVersity2 with 12 case studies with organizations as diverse as the National Cancer Institute, Federal Drug Administration, Department of Transportation, Office of the Bursar of the University of Maryland, or eBay. Our case studies demonstrated that TreeVersity2 is flexible enough to be used in different domains and provide useful insights for the data owners. A TreeVersity2 demo can be found at https://treeversity.cattlab.umd.edu.

Event sequence analysis is an important task in many domains: medical researchers may study the patterns of transfers within the hospital for quality control; transportation experts may study accident response logs to identify best practices. In many cases they deal with thousands of records. While previous research has focused on searching and browsing, overview tasks are often overlooked. We introduce a novel interactive visual overview of event sequences called \emph{LifeFlow}. LifeFlow is scalable, can summarize all possible sequences, and represents the temporal spacing of the events within sequences. Two case studies with healthcare and transportation domain experts are presented to illustrate the usefulness of LifeFlow. A user study with ten participants confirmed that after 15 minutes of training novice users were able to rapidly answer questions about the prevalence and temporal characteristics of sequences, find anomalies, and gain significant insight from the data.