Date & Time : Wednesday, October 16 06:15 pm - 07:30 pm
Location : TBD
Posters :
A Personal Life Review Photo Presentation Using Storyline Visualization Technique
Authors:
Yuka Nomura, Takayuki Itoh, Satoshi Nakamura
Abstract:
People often take photos as memorabilia by capturing moments and people who
are together. Since photo browsing is an effective way to reminisce the
memories, people like to use photo collections as enjoyable presentations for
themselves or their friends. In addition, personal photo presentation is
often used for psychological therapies. In this poster, we present a novel
technique for reviewing personal lives and reminiscing memories through the
presentation of personal photos. The technique first applies time-varying
clustering to organize a person-tagged photo collection, and then it employs
the storyline visualization method to layout the clustering result. It
further divides photo collections into two levels applying a hierarchal
clustering algorithm and chooses important photos as representative ones of
each cluster by solving the knapsack problem. After these pre-processing
steps, the photo collection is ready for browsing through a GUI with
interactive level-of-detail control. Our technique effectively helps user to
understand what kind of events and persons exist in each aspect of the photo
owner's life by exploring the personal photo collection.
A Visual Analytics System for Stock Data
Authors:
Shenghui Cheng, Zhifang Jiang, Zhiyuan Zhang, Klaus Mueller
Abstract:
Despite the occasional downfall, stocks have proven to be good investments.
However, making money in stock requires a great deal of research. With just
under 7,000 stocks listed in the US alone, picking the right stock at the
right time can be a daunting task. There are many metrics by which to rate
the prospects of a given stock and seasoned investors typically weight these
by their own experience to decide on an acquisition. In this poster we
describe a visual analytics system purposed to aid in this task.
A Visual Language to Characterise Transitions in Narrative Visualization
Authors:
Donia Badawood, Jo Wood
Abstract:
We use a taxonomy of panel-to-panel transitions in comics, refined the
definition of its components to reflect the nature of data-stories in
information visualization, and then, use the taxonomy in coding a number of
VAST challenges videos from the last four years. We represent the use of
transitions in each video graphically with a diagram that shows how the
information was added incrementally in order to tell a story that answers a
particular question. A number of issues have been taken into account when
coding transitions in each video as well as in designing and creating the
visual diagram such as, nested transitions, the use of sub-topics, and
delayed transitions.
A Visual Survey of Arc Diagrams
Authors:
Till Nagel, Erik Duval
Abstract:
Surveys are a common way of providing an overview over a family of
visualization techniques. In this poster we focused on arc diagrams, which
are an established method to visualize relations between nodes in a simple
path graph, and are laid out in one dimension. We collected a wide range of
examples of arc diagrams with different characteristics. Following
Jurgensmann and Schulz's poster on tree visualizations we present our
collection as a visual survey. As a result, our poster acts as visual
reference and as inspirational source.
Adaptive Grid-Like Layout
Authors:
Steve Kieffer, Tim Dwyer, Kim Marriott, Michael Wybrow
Abstract:
We explore various techniques to incorporate grid-like layout conventions
into a force-directed, constraint-based graph layout framework. In doing so
we are able to provide high-quality layout---with predominantly axis-aligned
edges---that is more flexible than previous grid-like layout methods and
which can capture layout conventions in notations such as SBGN (Systems
Biology Graphical Notation). Furthermore, the layout is easily able to
respect user-defined constraints and adapt to interaction in online systems
and diagram editors such as Dunnart. Interactive use of our techniques is
demonstrated at \\url{http://youtu.be/b7nV8dvAlIc}.
Architectural Patterns for Real-Time Visual Analytics on Streaming Data
Authors:
Eser Kandogan, Danny Soroker, Steven Rohall, Peter Bak, Frank van Ham, Jie Lu, Harold-Jeffrey Ship, Chun-Fu Wang, Jennifer Lai
Abstract:
Monitoring and analysis of streaming data, such as social media, sensors, and
news feeds, has become increasingly important. To effectively support
monitoring and analysis, statistical and visual analytics techniques need to
be seamlessly integrated; analytic techniques for a variety of data types
(e.g., text, numerical) and scope (e.g., incremental, rolling-window, and
global) must be properly accommodated; interaction and coordination among
several visualizations must be supported in an efficient manner; and the
system should support the use of different analytics techniques in a
pluggable and collaborative manner. In this poster we discuss architectural
patterns for real-time visual analytics based on building a real-time Twitter
monitoring application.
Bubble Heap Graphs
Authors:
Gi-nam Kim, Hyoji Ha, Byung-Won On, Kyungwon Lee, Manjai Lee
Abstract:
In this paper, we propose bubble heap graphs. In general, social graphs show
the overall relationship among nodes. For instance, from UN voting records, a
social graph can be drawn, where each node stands for a UN member (e.g., US)
and further two countries show similar voting patterns if the nodes
corresponding to the two countries are linked each other in the graph. In
such a social graph, we can clearly figure out the overall voting patterns of
all UN members. However, we often focus on a particular node in the graph. It
is plausible that a US citizen may have an interest in only US and he or she
wants to take a look at relationships between US and each of UN members. In
this case, using existing social graphs, it is hard to understand hidden
insight between US and any other country. In our problem, given a country
(called C) in which a user is interested, our proposed bubble heap graph
effectively visualizes the relationship between C and each of UN members. For
this bubble heap graph visualization, we present how to compute the
similarity value between nodes, and how to visualize the bubble heap graph.
In particular, to prove that our proposal is general-purpose, we applied our
visualization technique to two different data sets . (1) Voting records in UN
General Assembly and (2) Roll call data regarding US Senate sessions.
Cache Enhancement Methods for Out-Of-Core Pathline Computation
Authors:
Joong-Youn Lee, Jinah Park
Abstract:
Various visualization methods based on out-of-core technology have been
introduced to deal with large datasets whose sizes exceed the capacity of
local memory on a single desktop PC. However, the visualization of pathlines
for huge, time-varying datasets on a single machine is not feasible with
naive methods due to the low hit ratio of a general cache system based on
LRU. In this paper, we propose a novel cache management strategy which relies
on the processes of data pre-fetch and cache evacuation. We also evaluate the
proposed method using a real-world flow dataset. Our experiments show that
the proposed pre-fetch method increases the cache hit ratio whereas an
evacuation strategy decreases the re-fetch ratio.
Can Physical Visualizations Support Analytical Tasks?
Authors:
Simon Stusak, Aurelien Tabard, Andreas Butz
Abstract:
While physical objects have been used to represent information for a long
time, physical visualizations only recently started to attract attention from
the InfoVis and HCI communities. In this article we present our early
experiments in designing physical visualizations for supporting data
analysis. Based on Amar's taxonomy of analytical tasks we show that physical
visualizations can support a number of analytical activities but that further
research is needed to support all activities. Based on our analysis, we
propose promising directions for future research.
CateDocs: A New Visual Analytics Approach to Exploring High Dimensional Categorical Datasets
Authors:
Yueqi Hu, Xiaoke Huang, Chong Zhang, Yingyu Wu, Jing Yang, Ye Zhao, Scott Barlowe, Wei Chen
Abstract:
Most existing multidimensional visualization techniques do not work well for
high dimensional categorical datasets. The major challenges include
preserving the discrete nature of the data and visually exploring the high
dimensional space. In this poster, we propose a new visual analytics approach
for high dimensional categorical data. Our methodology is to convert a
categorical dataset into a document corpus and then apply advanced document
analysis and visualization techniques to the corpus. Two prominent knowledge
discovery tasks, namely cluster analysis and multivariate analysis, are
supported. For cluster analysis, the Latent Dirichlet Allocation (LDA) topic
model is employed to discover subspace clusters in a categorical dataset. The
clusters are then visualized in a semantically rich visualization for
interactive visual analysis. For multivariate analysis, LDA is used for
dimension reduction and optimal rule mining is used to discover rules
describing multivariate relationships in the reduced subspaces. The
effectiveness of this approach has been illustrated by case studies on real
datasets.
CiteVis: Exploring Conference Paper Citation Data Visually
Authors:
John Stasko, Jaegul Choo, Yi Han, Mengdie Hu, Hannah Pileggi, Ramik Sadana, Charles D. Stolper
Abstract:
Citation counts and intra-conference citations are one useful measure of the
impact of prior research in a field. We have developed CiteVis, a
visualization system for portraying citation data about the IEEE InfoVis
Conference and its papers. Rather than use a node-link network visualization,
we employ an attribute-based layout along with interaction to foster
exploration and knowledge discovery.
Classifying Visual Knowledge Representations: 23 Years On
Authors:
Francis C.B. Williams, Jonathan C. Roberts
Abstract:
We present a contemporary classification of visual knowledge representations
into clusters within a two dimensional space, as well as a visual hierarchy
depicting pairwise similarity between representations. The work in this paper
follows the process carried out by Lohse et al. in their paper entitled
`Classifying Visual Knowledge Representations' (Lohse et al. 1990). The
images for the study were collected from the VisWeek 2012 proceedings for the
InfoVis conference. The results show a progressive blending of visualization
types that was not so prevalent in the Lohse paper. This is indicative of the
progression from distinct visualization categories to composite designs
within the visualization community.
CorrelatedMultiples: Spatially Coherent Small Multiples with Constrained Multidimensional Scaling
Authors:
Xiaotong Liu, Yifan Hu, Stephen North, Teng-Yok Lee, Han-Wei Shen
Abstract:
Small multiples are a popular method of summarizing and comparing multiple
facets of complex data sets. Since they typically do not take into account
correlations between items, serial inspection is needed to search and compare
items, which can be ineffective. To address this, we introduce
CorrelatedMultiples, an alternative of small multiples in which items are
placed so that distances reflect dissimilarities. We propose a constrained
multidimensional scaling (CMDS) solver that preserves spatial proximity while
forcing items to fit within a fixed region. We evaluate the effectiveness of
CorrelatedMultiples through a controlled user study, and compare the CMDS
method with competing methods. We also demonstrate the usefulness of
CorrelatedMultiples in a case study on visual analysis of stock market
trends.
DQvis: A Toolkit for Visual Quality Analysis for Relational Database
Authors:
Kun Wang, Dongxing Teng, Haiyan Yang, Cuixia Ma, Hongan Wang
Abstract:
Data quality plays an important role in information systems visual analysis
is an effective and efficient method of data quality evaluation. As most of
the existing researches of visual data quality analysis are aimed at specific
kinds of problems instead of an overall solution, we propose a relatively
complete solution for data quality analysis of relational database, including
the seldom researched normal form analysis. Then we built a set of
interactive visual forms and finally develop a toolkit DQvis to facilitate
data quality analysis.
DreamVis: Visualizing Logged Dream Data
Authors:
Riane Vardeleon, Sheelagh Carpendale
Abstract: We take a look at information visualization from a personal standpoint, creating a series of visualizations to explore the possibilities of more accessible representations of one's dreams. DreamVis is designed to make it more possible for people to examine and reflect on their dreams through the use of visuals. Most, if not almost all, dreams are forgotten within the first few moments upon waking. With DreamVis we open the set of personal visualizations, which have more commonly been used to visualize more concrete activities such as diet and exercise, to include more abstract and emotional data set within the context of their daily life.
Enabling Visual Exploration of Long-term Physiological Data
Authors:
Miriam Zisook, Javier Hernandez, Matthew S. Goodwin, Rosalind W. Picard
Abstract:
With the recent development of wearable and comfortable biosensors, larger
datasets of physiological data are being collected in challenging real-life
scenarios. In order to gain insight from these datasets, behavioral
scientists need tools that enable agile and efficient data exploration. In
this work, we designed and implemented two visualization tools for
large-scale time-based datasets, and combined the lessons learned to create a
more general-purpose tool.
Error Bars Considered Harmful
Authors:
Michael A. Correll, Michael Gleicher
Abstract:
Confidence intervals, standard error, and generalized error rates are
typically visualized with error bars - thin strokes that are superimposed
over the mean. This work present the results of crowd-sourced experiments
which illustrate that viewers misinterpret these encodings even at the most
basic level (where one would hope larger margins of error reduce the
confidence in judgments about means). We then present evaluations of three
alternate (or supplemental) visual encodings for the same task and show that
choice of visual encoding can result in viewers who make decisions which are
better informed by the margins of error.
Exploring Interaction Techniques and Task Types for Direct-Touch as Input Modality
Authors:
Matthias Nielsen, Mikkel Baun Kjaergaard, Kaj Gronbaek
Abstract:
Recent InfoVis interaction research has called for the inclusion of novel
interaction technologies for InfoVis tools. However, it remains an
underexplored question what impact novel interaction technologies will have
on the existing body of work addressing interaction techniques and task
types. This study presents results based on a prototype of an InfoVis tool
developed to substitute conventional mouse input with direct-touch as input
modality. Two versions of the prototype will be presented: one geared towards
mouse input and one geared towards direct-touch input, and the implications
entailed by introducing direct-touch as an input modality will be discussed.
Exploring Subjective Survey Classification of a Photographic Archive using Visual Analytics
Authors:
Alexander Kachkaev, Jo Wood
Abstract:
We use an interactive visual analytics approach to explore the results of a
survey and utilise parallel coordinate plots and small multiples as key
visualization techniques. The scope of the survey is a set of 900 photographs
from 3 origins, which were to be subjectively classified by the participants
in a number of ways. In this poster we describe the interface of the designed
tool and also highlight the findings it allowed us to make. By visualizing
the collected survey data and navigating through it we could estimate the
proportions of different types of photographs, identify qualitative
differences between their sources and correlate the responses with image
metadata. We were also able to support the survey process itself: with
various visual representations of the collected results we could detect
inappropriate behaviour of a number of participants, handle issues related to
unavailability of some photographs and also ensure responses sampled the
image database appropriately.
Exploring the Effect of Visualisation Structure on a Text Analysis Task
Authors:
Adam Duncan, Chris Baber
Abstract:
In this research we explore how analyst-constructed visualisations of themes
in document collections can act as both a communicative product and an
exploratory, insight-generating process. Our aim is to employ visual
storytelling to create new visualisation and diagramming techniques for
summarising information from a variety of sources. Presented here are the
results of a pilot study which compares two visualisation tools, one based on
storyboarding, and the other based on Concept Mapping. We contribute a
storyboarding visualisation that is novel in the context of visual analytics
and intelligence analysis, and is built around the concept of storytelling.
The results from our study imply that it performs similarly to a more
established technique like Concept Mapping in terms of the relevance of
information recorded, but it was preferred by participants when they were
required to infer meaning from the visualisations.
GapFlow: Visualizing Gaps in Care for Medical Treatment Plans
Authors:
David Gotz, Nan Cao, Esther Goldbraich, Boaz Carmeli
Abstract:
Despite the widespread adoption of clinical guidelines (i.e. canonical
treatment plan templates that represent generally accepted best practices),
significant variations in care are often found across a population of
patients. Gaps between the actual treatment programs performed on patients
and the recommended guidelines are inevitable given the complexity of
disease, differences between patients, and the individualized
patient-centered decisions made by clinicians during each encounter. This
poster presents a visualization tool designed to help clinical organizations
better understand these gaps in care. We describe the input data, our
analysis technique to classify individual gap events, and an interactive
visualization technique which aggregates and summarizes the results for
clinical interpretation.
Graph-Based Navigation of a Box Office Prediction System
Authors:
Mat Kelly, Michael L. Nelson, Michele C. Weigle
Abstract: Predicting movies' box office performance is a bit like gambling and predicting the stock market: there is a lot of luck involved but analyzing the system's input frequently leads a better result. There are many inputs to box office prediction but certainly among the most influential in predicting future success are the actors involved and their past performances. Grasping the variables needed in the large data sets to make a box office prediction is an overwhelming task where a visualization would be useful. In this paper we describe a means to visualize these large data sets and distill them down into a simple causal approach that makes understanding the factors that affect box office performance easier.
Hierarchical Qualitative Color Palettes
Authors:
Martijn Tennekes, Edwin de Jonge
Abstract:
Color is an important means to display categorical data in statistical
graphics. Categories are often hierarchically structured in a classification
tree, but most qualitative color palettes do not take this hierarchy into
account. We present a method to map tree structures to colors from the
Hue-Chroma-Luminance (HCL) color model. The HCL color space is known for its
well balanced perceptual properties. Our study suggest that hierarchical
qualitative color palettes are very useful: not only for improving standard
hierarchical visualizations such as trees and treemaps, but also for showing
tree structure in non-hierarchical visualizations.
Hyperslice Visualization of Metamodels for Manufacturing Processes
Authors:
Sascha Gebhardt, Bernd Hentschel, Torsten Kuhlen, Toufik Al Khawli, Wolfgang Schulz
Abstract:
In modeling and simulation of manufacturing processes, complex models are
used to examine and understand the behavior and properties of the product or
process. To save computation time, global approximation models, often
referred to as metamodels, serve as surrogates for the original complex
models. Such metamodels are difficult to interpret, because they usually have
multi\-dimensional input and output domains. We propose a hyperslice\-based
visualization approach, that uses hyperslices in combination with direct
volume rendering, training point visualization, and gradient trajectory
navigation, that helps in understanding such metamodels. Great care was taken
to provide a high level of interactivity for the exploration of the data
space.
Identifying Risk Factors for Birth Defects in High Dimensional Environmental Health Data
Authors:
Chong Zhang, Jing Yang, F. Benjamin Zhan, Xi Gong, Jean D. Brender, Peter Langlois, Scott Barlowe
Abstract:
Scientists are connecting massive environmental pollution data with birth
registries to identify birth defect risk factors. It is a significant
challenge to identify associations between maternal exposure to toxic
chemicals and malformations in offspring in this type of study. We propose a
novel visual analytics approach to addressing this challenge. The approach
consists of a visual analytics pipeline where analysts can gradually and
interactively refine the set of potential risk factors. The approach tightly
integrates the following techniques: (1) Statistical analysis methods such as
Point-Biserial Correlation analysis and Logistic Regression; (2) a risk
pattern mining technique; and (3) visualization techniques such as Parallel
Coordinates and the Value and Relation display. We demonstrate the usefulness
of the approach using a dataset from an ongoing case-control study in Texas.
Illustrative Data Graphics in 18th&19th Century Style: A Case Study
Authors:
Benjamin Bach, Pierre Dragicevic, Samuel Huron, Petra Isenberg, Yvonne Jansen, Charles Perin, Andre Spritzer, Romain Vuillemot, Wesley Willett, Tobias Isenberg
Abstract:
We present a scientific reflection on the creation of a large illustrative
personal data graphics that we create to mimic 18th--19th century style
visualizations. We discuss the relationship to illustrative visualization,
the story-telling aspects of the design, the used data sources and their
purpose, the creation of the overall style, and the aspects of the
multi-author creation process.
Image-Based Exploration of Iso-Surfaces for Large Multi-Variable Datasets using Parameter Space
Authors:
Roba Binyahib, Madhusudhanan Srinivasan, Christopher Knox
Abstract:
With rapid advances in HPC resources, more complex simulations have resulted
in larger data size, with higher resolution and many variables. Visualizing
large multivariate datasets is a challenging problem that often requires
high-end clusters. Consequently, novel visualization techniques are needed to
explore such complex data. Explorable image (EI) is a novel approach that
provides limited interactive visualization without the need to rerender from
the original data. In this work, we used the concept of EI to create a
workflow that generates explorable iso-surfaces for scalar fields in a
multivariate, time-varying dataset. We present a run-time tool that allows
the user to interactively browse and calculate a combination of iso-surfaces
superimposed on each other. The result is the same as calculating multiple
iso-surfaces from the original data but without the memory and processing
overhead. Our tool also allows the user to change the (scalar) values
superimposed on each of the surfaces, modify their color map, and
interactively re-light the surfaces. We demonstrate the effectiveness of our
approach over a multi-terabyte combustion dataset. We also illustrate the
efficiency and accuracy of our technique by comparing our results with those
from a more traditional visualisation pipelines.
Interacting with Data Visualizations on Tablets and Phones: Developing Effective Touch-based Gestures and Operations
Authors:
Ramik Sadana, John Stasko
Abstract:
Currently, data visualization systems employ a variety of techniques and user
interface components to facilitate interaction on mouse-based desktop and
laptop computers. However, the same is not true for interaction on
touch-based devices such as tablets and mobile phones. Data visualization has
yet to become pervasive there and it is not clear what interactive operations
should be provided and how they should be performed. In this paper, we
discuss an approach to designing interactions for touch- based visualizations
employing user-suggested gestures. This design space for touch gestures on
data visualizations is especially rich, perhaps more so than many expect.
Interactive 3D Gaze Visualization for Contiguous Cross-sectional Medical Images
Authors:
Hyunjoo Song, Jihye Yun, Bohyoung Kim, Jinwook Seo
Abstract:
Gaze visualization has been used in the medical field to understand how
radiologists read medical images. While prior works were mainly based on
diagnoses with a single image, recent works focus on diagnoses with
consecutive cross-sectional medical images acquired from preoperative
computed tomography (CT) or magnetic resonance imaging (MRI). Such images
have distinct characteristics that hundreds of them are from a single exam
composing a natural 3D spatial structure. Radiologists have to scroll through
a stack of the images for a diagnosis, resulting in more complicated gaze
patterns to visualize. Little work has been done on visualizing such gaze
patterns for contiguous cross-sectional medical images. We present an
interactive 3D gaze visualization, where InfoVis and SciVis techniques are
harmonized to show the abstract gaze data along with a realistic 3D rendering
of the visual stimuli (i.e. organs and lesions).
Interactive Multi-resolution Exploration of Million Node Graphs
Authors:
Zhiyuan Lin, Nan Cao, Hanghang Tong, Fei Wang, U Kang, Duen Horng Chau
Abstract:
We are working on a scalable, interactive graph visualization system to
support multi-resolution exploration of million-node graphs in real time. By
adapting a state-of-the-art graph algorithm, our prototype system generates a
multi-resolution view of graphs with up to 68 million edges under a few
seconds. We are experimenting with interaction techniques that help users
interactively explore this overview and drill down into details. While many
visualization systems for million-node graphs require dedicated servers to
process the graphs, our prototype runs on a commodity laptop computer. We aim
to handle graphs that are at least an order of magnitude (100M edges) larger
than what current systems can support.
Interactive Visualization of Brain Volume Changes
Authors:
Claudia Hanel, Bernd Hentschel, Torsten Kuhlen, Peter Pieperhoff, Katrin Amunts
Abstract:
The visual analysis of brain volume data by neuroscientists is commonly done
in 2D coronal, sagittal and transversal views, limiting the visualization
domain from potentially three to two dimensions. This is done to avoid
occlusion and thus gain necessary context information. In contrast, this work
intends to benefit from all spatial information that can help to understand
the original data. Example data of a patient with brain degeneration are used
to demonstrate how to enrich 2D with 3D data. To this end, two approaches are
presented. First, a conventional 2D section in combination with transparent
brain anatomy is used. Second, the principle of importance-driven volume
rendering is adapted to allow a direct line-of-sight to relevant structures
by means of a frustum-like cutout.
Interpolation of Non-Gaussian Probability Distributions for Ensemble Visualization
Authors:
Brad E. Hollister, Dr. Alex Pang
Abstract:
A typical assumption is that ensemble data at each spatial location follows a
Gaussian distribution. We investigate the consequences of that assumption
when distributions are non-Gaussian. A sufficiently acceptable interpolation
scheme needs to be addressed for the interpolation of non-Gaussian
distributions. We present two methods to calculate interpolations between two
arbitrary distributions and compare them against two baseline methods. The
first method uses a Gaussian Mixture Model (GMM) to represent distributions.
The second method is a non-parametric approach that interpolates between
quantiles in the cumulative distribution functions. The baseline methods for
comparison purposes are: (a) using a Gaussian representation and
interpolating the means and standard deviations, and (b) forming a new
distribution based on the interpolation of individual realizations of the
ensemble. We show that the two proposed non-Gaussian interpolation methods
have the following behavior: the interpolated distributions do not decompose
to more constituent Gaussian distributions than the highest modality of those
being interpolated, and do not have variances less than the smallest variance
from the grid points being interpolated. Finally,we compare these four
interpolation methods when used in the analysis of scalar and vector fields
of ensemble data sets, particularly in areas where the distribution is
non-Gaussian.
Lightness Constancy in Surface Visualization
Authors:
Danielle Albers, Alper Sarikaya, Michael Gleicher
Abstract:
Color is an effective and commonly used channel for displaying data in
surface visualization. However, color is affected by shadows and shading,
which convey depth and shape on surfaces. The human visual system has evolved
constancy mechanisms for identifying color under varying illumination
conditions. Lightness constancy allows people to accurately perceive colors
in real shadows; however, its effectiveness in surface visualizations is not
well understood. We report a series of initial studies that confirm the
existence of lightness constancy effects on molecular surface renderings. We
evaluate common design decisions to show how choices of attenuation, color
ramp, and shading model impact viewers' abilities to accurately identify
colors on molecular surfaces rendered with ambient occlusion. Our findings
lead to a set of generalizable design implications for effective surface
visualization using color.
Modeling Incremental Visual Analytics Visualizations
Authors:
Marco Angelini, Giuseppe Santucci
Abstract:
This work presents a formal model for characterizing the iterative drawing of
a visualization, describing the practical issues and outlining the main
parameters that can be used to drive and evaluate the whole process. The
model is validated against an incremental Visual Analytics application.
Modeling of Clouds from Weather Forecast Data
Authors:
Guang Yang, Chunqiang Yuan, Shiyu Hao, Xiaohui Liang
Abstract:
Cloud visualization of weather forecast data plays an important role in
atmosphere analysis and weather prediction. However, due to the low
resolution of weather forecast data and complicated geometry of clouds, it is
a challenging problem to present realistic three dimensional clouds. In this
paper, we propose a new framework to solve this problem. Firstly, data points
are classified into three types of clouds: cumulus, stratus and cirrus. Then,
an interpolation method is used to construct large-scale stratus. For cirrus,
we apply a searching process. Finally, clouds are rendered in interactive way
with various view directions. Experimental results demonstrate that the
proposed method can yield large-scale and realistic clouds.
Modeling User Interactions for Complex Visual Search Tasks
Authors:
J. Helen Zhao, Quan Lin, Alvitta Ottley, Remco Chang
Abstract:
Modeling a user's interactions is intimately tied to many areas of research
in the fields of HCI and Visual Analytics. Most notably, developing adaptive
visual interfaces and effectively prefetching for large datasets, first
requires understanding the user's behavior and analytical process. In this
work, we demonstrate the potential of using a user's mouse movements and
clicks to achieve this goal. In an online study, we gather users'
interactions as they perform a complex visual search task. Our results
indicate a significant difference between the search strategies employed by
users who were quick at completing the task and those who were slow.
Molecular Trajectory Projections using Molecular Symmetry
Authors:
Kyle Wm. Hall, Peter G. Kusalik, Sheelagh Carpendale
Abstract:
We herein present Radially, Angularly Mapped Trajectory (RAMT) plots. RAMT
plots are trajectory traces plotted in terms of angles and radii in
projections of 3D simulations based on the underlying symmetry of a chemical
species of interest. Trajectory traces can represent average properties and
instantaneous configurations in a single visualization. We explain and
illustrate the RAMT plot technique in the context of molecular dynamics
simulations of two hydroxyl radicals in a solvent.
Multimedia Pivot Tables
Authors:
Marcel Worring, Dennis C. Koelma
Abstract:
Image collections are a tremendous source of information. Yet due to the
semantic gap it is difficult to get access to their content, while at the
same time it is difficult to properly employ their context such as tags and
metadata. To move forward we propose a multimedia analytics solution. The
most widespread and universally used analytic tools are spreadsheets, where a
powerful feature is the possibility to generate pivot table reports. They
provide flexible interactive summaries of the data along various dimensions.
Pivot tables have been designed and are in use for structured data. Our goal
is creating pivot tables for accessing collections of images, their content,
tags, and metadata. This is a challenging task as automatic descriptors for
image content are noisy, tags are numerous and subjective, and metadata can
have many types. To tackle these challenges we present methods and
visualizations for semi-interactively categorizing an image collection and
from there design and develop pivot tables for such a collection.
PetroVis: Exploratory Visualization for Petrographic Characterization
Authors:
Ahmed E. Mostafa, Juliana Cevolani, Emilio Vital Brazil, Ehud Sharlin, Mario Costa Sousa
Abstract:
The process of petrographic analysis aims to help oil and gas experts in
building a prediction model which would better characterize and explain the
behavior of the hydrocarbon reservoir. However, making informed decisions
regarding a good characterization requires exploration and integration of
very large amounts of data presenting many challenges in which high
dimensionality is the most important one. We present PetroVis, our
visualization prototype which we developed with the goal of supporting the
visual analysis of petrographic data. PetroVis incorporates the use of
multidimensional visualization techniques coupled with statistical methods in
order to enable identification, classification, validation and interpretation
of petrofacies. We developed PetroVis continuously guided by the feedback and
suggestions from our domain collaborator.
Reconstruction of Sharp Features from Industrial CT Data
Authors:
Arindam Bhattacharya, Rephael Wenger
Abstract: We describe an algorithm ReliableGrad to construct reliable gradients from scalar data with sharp features. We combine ReliableGrad with our previously published algorithm MergeSharp to construct isosuraces with sharp edges and corners from industrial CT data. The resulting isosurface meshes have sharp edges and corners reliably represented by mesh edges and vertices.
Recording Reusable and Guided Analytics From Interaction Histories
Abstract:
The use of visual analytics tools has gained popularity in various domains,
helping users discover meaningful information from complex and large data
sets. Users often face difficulty in disseminating the knowledge discovered
without clear recall of their exploration paths and analysis processes. We
introduce a visual analysis tool that allows analysts to record reusable and
guided analytics from their interaction logs. To capture the analysis
process, we use a decision tree whose node embeds visualizations and guide to
define a visual analysis task. The tool enables analysts to formalize
analysis strategies, build best practices, and guide novices through
systematic workflows
Rendering Point Clouds with Feature Textures
Authors:
Yuping Zhang, Marc Olano, Jonathan P. Dandois, Jian Chen
Abstract:
3D point cloud rendering is an efficient way to represent 3D scenes. For
forest ecologists, 3D point cloud models are also useful for measuring
attributes of forest canopy structure and color at the scale of individual
tree crowns. Rendering forest point clouds in the web browser may also help
scientists cooperate and use data efficiently. The main problem of point
cloud rendering is it lacks details without an extremely dense cloud, which
may not be easy to acquire. Our approach is to render point clouds
constructed by computer vision structure-from-motion feature matching and
using textures from the feature points in the original images. These textures
are rendered on patches, oriented to match the original camera orientation.
User study shows this method provides more details and can help ecologists
with their research.
ScagExplorer: Using Scagnostics to Cluster Huge Datasets
Authors:
Tuan Nhon Dang, Leland Wilkinson
Abstract:
We introduce a method for guiding interactive exploration of high-dimensional
data. The method is based on nine characterizations of the 2D distributions
of orthogonal pairwise projections on a set of points in multidimensional
Euclidean space. These characterizations include measures such as, density,
skewness, shape, outliers, and texture. Using with these measures, we can
quickly generate a comprehensive summary of the 2D relations of variables in
a large dataset with more than a hundred dimensions.
Serendip: Turning Topics Back to the Text
Authors:
Eric Alexander, Joe Kohlmann, Robin Valenza, Michael Gleicher
Abstract:
Statistical topic modeling is an increasingly popular approach to text
analysis. Many existing visualization tools focus on analyzing the model
itself, distinct from the documents upon which it was trained. In contrast,
we seek to treat the model as a lens through which to view the original
documents. This would enable the reader to observe trends and build
hypotheses at multiple scales--ranging from across a corpus to within a
single text--and find both algorithmic data and textual examples to defend
these hypotheses. Supporting this workflow requires a multi-tiered framework
that affords comparisons at three levels: the entire corpus, small sets of
documents, and a single document. This framework is embodied in Serendip, a
web-application that combines view-coordinated reorderable matrices, small
multiples displays, and tagged text in order to allow readers to develop
insight at and across multiple levels.
Size Judgment and Comparison in Tag Clouds
Authors:
Khaldoon (Kal) Dhou, Robert Kosara, Mirsad Hadzikadic, Mark Faust
Abstract:
Tag clouds can be used for a variety of purposes, like providing a high-level
understanding of a document. It is still unclear how users perceive the font
size of the words in tag clouds and how they make their judgments of the size
of words. In this poster, we look at how users estimate the relative sizes of
words given different characteristics. We studied the influence of
decorations like filled areas, boxes, and shadows to determine whether they
would influence the perceived size. Another parameter we tested was the
appearance of words (i.e. by choosing words with and without ascenders and
descenders). We found significant effects from all of those parameters, which
suggests that designers of tag clouds need to be aware of the influence of
design choices on the perceived data.
StFT-Stereoscopic Filtering Technique
Authors:
Ragaad AlTarawneh, Achim Ebert, Eduard Kosel
Abstract:
In this work, we propose a new technique, called the Stereoscopic Filtering
Technique (StFT), based on the stereoscopic highlighting approach. The
proposed technique utilizes the depth cue to filter the large graph
structures. One of the main abilities of our technique is that it isolates
the specific graph portions to magnify them for a detailed exploration
without affecting other visual cues such as color or shape. It offers a
natural way to provide focus+context technique as it brings some parts of the
graph closer to the viewer while keeping the rest in the background. We use
this technique to query about specific topological features about the graph
such as trees, clusters, complete graphs, or unknown graph structures. The
preliminary results show the intuitiveness of the technique's approach.
Moreover, they indicate the opportunities to handle the scalability issues in
querying 2D representations for large graphs.
The Natural Materials Browser: Using a Tablet Interface for Exploring Volumetric Materials Science Datasets
Authors:
Angus Graeme Forbes, Tony Fast, Tobias Hollerer
Abstract:
We present a novel tablet application, the Natural Materials Browser, that
allows a user to interact with volumetric datasets created from a series of
natural materials samples. The data samples -- high resolution meso-scale
volumetric images of nutshells gathered via micro-computed tomography -- are
envisioned as virtual specimens presented many orders of magnitude larger
than their characteristic length scale. The user, initially placed in the
center of the volumetric dataset and facing orthogonally toward the original
2D image slices, uses an iPad tablet as a magic lens to view and navigate the
data via physical rotation and multitouch gestures. The user has simultaneous
access to multiple representations of the datasets from any angle or
position, and an additional viewport provides real-time, spatial statistics
on the current view of the currently loaded dataset. We conducted a
preliminary evaluation of the application by collating cognitive walkthroughs
given to domain experts in materials science. Their feedback indicated that
our tablet application could potentially be an effective tool for enabling
insights regarding these data samples and, more generally, that it functions
as a low-cost, immersive system with which to explore volumetric datasets.
Time Ring Maps: Visualization for Spatiotemporal Sensor Data
Authors:
P. Bak, J. von Kaenel, X. Sun, I. Dumitrescu, C. Vecchiola, O. Amarilio
Abstract:
We propose a method that can visualize the spatiotemporal measurements
originating from a sensor network. Each sensor records a sequence of
measurements and the accumulation of these measurements is of central
interest. Importantly, the sensor can associate with the attribute that
contains a hierarchy on its values. Our method, called Time Ring Maps, can 1)
capture the sequence and distribution of the measurements, 2) allow the user
to compare temporal patterns between sensors, and 3) display the temporal
patterns at different hierarchy levels when users zoom the background map.
The applicability of the Time Ring Maps is demonstrated on real world data
from traffic network monitoring.
Time Series Modeling for Smart Grid Monitoring
Authors:
Georg Fuchs, Natalia Andrienko, Gennady Andrienko
Abstract:
Analysis and modeling of time series (TS) data plays an important role in
monitoring and control tasks related to large and/or complex systems, such as
electric power grids. Typically, such time series exhibit nested temporal
cycles (e.g., hours, days, weeks) inherent to human activities. We here
present a TS modeling approach that makes explicit use of this inherent
cyclicity for the purpose of providing appropriate prediction of
time-dependent parameters for situation assessment and decision support. The
approach is part of ongoing research towards providing Visual Analytics
support in the context of critical infrastructure monitoring.
Time-Order Kinetic Irreversible Compression Scheme for Visualization of Large Particle System
Authors:
Hiroaki Ohtani, Katsumi Hagita, Atsushi M. Ito, Tsunehiko Kato, Takayuki Saitoh, Takaaki Takeda
Abstract: We propose in this paper a data compression scheme for large-scale particle simulations, which has favorable prospects for scientific visualization of particle systems. Our scheme deals with a time sequence of particle data obtained from simulations as a set of individual particle trajectories and compresses it by approximating each trajectory as a piecewise polynomial function that satisfies a given tolerance; for each approximated trajectory, a piecewise polynomial is determined so that the deviations from the original trajectory are smaller than the tolerance at all data points. The approximated trajectories are continuous within the entire time interval and therefore the particle position can be evaluated everywhere in the interval. The scheme is also capable of using an independent time-step for each particle. We name this concept TOKI (Time-Order, Kinetic, and Irreversible) compression. In this paper, we show an application result in the plasma particle simulation data by the data-compression scheme under this concept.
Towards a Characterization of Guidance in Visualization
Authors:
Hans-Jorg Schulz, Marc Streit, Thorsten May, Christian Tominski
Abstract:
Applying and parameterizing advanced visualization tools for solving
different problems can be difficult for users, who are not necessarily
visualization experts. The visualization community has begun to address this
problem by developing assistive approaches under different labels. In this
work, we propose an initial version of a characterization scheme for guidance
in visualization. With the help of the characterization, the visualization
community will be able to categorize existing approaches and to identify
white spots, which are so far underrepresented and require more research.
Towards a Formalized Process for Creating Haptic Data Visualizations
Authors:
Panagiotis D. Ritsos, Sabrina A. Paneels, Peter J. Rodgers, Jonathan C. Roberts
Abstract:
Haptic Data Visualization (HDV) is a novel application of haptics. It
provides functionality by which users touch and feel data, making it a useful
tool for users with vision impairments. However, creating such visualizations
usually requires programming knowledge, that support workers and tutors of
blind users may not possess. To address this issue we propose a formalized
process for creating HDVs using the HITPROTO [5] toolkit, which requires no
programming experience. We further illustrate this process using an example
HDV.
UberShadie: A Domain-Specific Language for General Volume Processing and Visualization on Heterogeneous Parallel Systems
Authors:
Hyungseok Choi, Hanspeter Pfister, Won-Ki Jeong
Abstract:
In this paper, we introduce our on-going work on developing a domain-specific
language specifically designed for 3D data processing and visualization on
heterogeneous parallel computing systems. Our method is inspired by the
previous work by Hasan et al., (Shadie [5]). We observed that the language
design resembling Python is important for novice users, and the high-level
abstraction of GPU programming is another advantage of Shadie. However, we
also observed that the shader-like framework in Shadie becomes a major
obstacle that significantly impairs the flexibility of the system. Based on
these observations, we propose to develop a more advanced, flexible, and
easy-to-use programming language, compiler, and runtime system, called
UeberShadie, specifically designed to easily write a research code handling
3D volume data while providing superior computing performance by leveraging
the state-of-the-art heterogeneous parallel computing technology, such as
multicore CPUs and GPUs (Graphics Processing Units).
Understanding Performance of Protein Structural Classifiers
Authors:
Alper Sarikaya, Danielle Albers, Michael Gleicher
Abstract:
Many bioinformatics applications utilize machine learning techniques to
create models for predicting which parts of proteins will bind to targets.
Understanding the results of these protein surface binding classifiers is
challenging, as the individual answers are embedded spatially on the surface
of the molecules, yet the performance needs to be understood over an entire
corpus of molecules. In this project, we introduce a multi-scale approach for
assessing the performance of these structural classifiers, providing
coordinated views for both corpus level overviews as well as
spatially-embedded results on the three-dimensional structures of proteins.
Urban Transport Energy Consumption Explored Through 3D Arc Maps
Authors:
Stephanie Schweitzer, Ariane Middel, Wenwen Zhang
Abstract:
We present a visualization tool for the analysis of the transport energy
budget of a city, including the life-cycle energy embedded in the built
infrastructure and the energy consumption by transportation. In our
application, cumulative total energy is displayed as stacked, color-coded
cylinders on a land use map. Transport energy consumption from traffic
between inner-city travel zones is visualized as directed 3D Bezier curves.
Our tool will support urban planners in assessing how urban form and
infrastructure impact energy consumption and related greenhouse gas (GHG)
emissions.
Using Eye-Tracking as Interactive Input Enhances Graph Visualization
Authors:
Mershack Okoe, Sayeed Safayet Alam, Radu Jianu
Abstract:
Much of the visual analysis of a graph reduces to a set of building-block
visual tasks such as node scanning or edge and path tracing. These tasks may
be trivial in small graphs but increase in complexity the larger and denser a
graph visualization becomes. In this work, we use eye-tracking data as a
real-time input to alter a graph visualization interactively and support its
analysis. Specifically, we display labels of fixated nodes, we highlight
edges as they are visually traced, and we dim out edges that pass through a
users view-focus while having their endpoints far outside of it. We conducted
a small informal user study to compare the performance of our eye-tracking
enabled graph visualization versus a graph visualization system that only
uses mouse input. The gaze-enabled visualization performed better in terms of
accuracy and response time and was preferred by all participants.
Using Pixel-Based Visualizations to Detect Adverse Drug Events
Authors:
Ming Hao, Sebastian Mittelstadt, Meichun Hsu, Umeshwar Dayal, Joseph Terdiman, Daniel A. Keim
Abstract:
Adverse reactions to drugs are a major public healthcare issue. Currently,
the Food and Drug Administration (FDA) publishes quarterly reports that
typically contain in the order of 200,000 adverse incidents. In such a large
number of incidents, low frequency events that may be highly clinically
significant but are often undetected. In this poster, we introduce a pixel
cell-based visualization technique with novel relevance ordering algorithm,
significance statistics computation, and semantic zooming in an x-y plan. We
are able to identify important adverse events, such as the known association
of the drug Avandia with myocardial infarction; as well as low frequency
events such as the association of Actos with bladder cancer.
Using Sparklines to Reveal Trends in Parallel Coordinates
Authors:
Tomasz Opach, Jimmy Johansson, Jan Ketil Rod
Abstract:
In this poster we discuss the concept of sparklines that can be used for
analysing trends in cluttered displays in parallel coordinates. The parallel
coordinates sparklines have been implemented in the web-based visualization
tool ViewExposed to help users analyse data on exposure and vulnerability to
natural hazards. To validate the usability of our proposed method, we have
conducted user tests with 53 participants and our results show that the
sparklines are particularly useful for revealing significant trends.
Using Visual Analytics to Understand Social and Communicative Behaviors
Authors:
Yi Han, Agata Rozga, John T. Stasko, Gregory D. Abowd
Abstract:
Technologies are providing new opportunities for psychologists to record and
study human behaviors in unprecedented detail. In time, such rich behavioral
datasets will be collected by psychologists everywhere. We are studying the
use of technologies to capture, measure, analyze and understand human social
and communicative behaviors. However, the massive amount of
video/audio/sensing data collected during multimodal interactions from
hundreds of subjects can be difficult to explore for psychology researchers.
We investigate how visual analytics can be of help in providing a new method
for exploring such challenging datasets.
Visual Analysis of Circular Dichroism Spectra in Molecular Biophysics
Authors:
Daniel Engel, Christina Gillmann, Sebastian Fiedler, Sandro Keller, Inga Scheler, Hans Hagen, Christoph Garth
Abstract:
Among a wide range of applications, Circular Dichroism (CD) spectroscopy may
be used to investigate the conformational changes of proteins. We describe
the design of a visual framework for the analysis of data from automated CD
spectroscopy. Based on a requirement description for model-based analysis of
protein conformation changes, design choices for visual encodings and
interaction techniques are explored and related to existing work in a
different application. The results show how an entwined optimization and
visualization can help to glean insights in molecular biophysics and protein
chemistry.
Visual Analysis of Ionospheric Disturbance Hypotheses about Earthquake
Authors:
Fan Hong, Siming Chen, Hanqi Guo, Xiaoru Yuan, Jian Huang, Yongxian Zhang
Abstract:
In seismic research, a working hypothesis is that ionospheric disturbances is
related to lithosphere activities such as earthquakes. Domain scientists are
working to find patterns from certain ionospheric attributes related to
seismic activities. However, to find patterns from large amount of data is
challenging, since it is hard to extract, formulate and search patterns. To
address on these challenges,we developed an interactive system which supports
the workflow of seismic research on ionospheric data. Our system can assist
domain scientists to propose and examine hypotheses on relationships between
ionospheric disturbance and seismic activities.
Visual Analytics of Sentiment Trends in Social Media Streams: The 2013 Confederation Cup Case
Authors:
Maira Gatti, Alexandre Rademaker, Daniel Lemes, Paulo Cavalin, Claudio Pinhanez, Rogerio de Paula
Abstract:
Millions of people post messages every day in social media net- works,
especially on microblogging ones, like Twitter. There has been a major effort
on monitoring all those messages for social media analytics to boost social
media actions like marketing campaigns. Although there has been some
approaches to detect and visualize topics trends of social media text
analytics, this is an area full of challenges and open problems. We tackle in
this poster the problem of visualizing real-time topics trends with sentiment
analysis of streaming Twitter data from Brazilian users during games of the
2013 FIFA Confederations Cup. We compute the co-related matrix of terms
occurrence to reduce the original terms matrix sparsity and therefore to
select the most relevant topics associated to each player to be visualized
through time series.
Visual Trend Analysis in Weather Forecast
Authors:
Alexandra Diehl, Stefan Bruckner, M. Eduard Groller, Claudio Delrieux, Celeste Saulo
Abstract:
In this work, we propose an interactive approach for visual analysis of
weather trends and forecast errors in short-term weather forecast
simulations. Our solution consists of a multi-aspect system that provides
different methods to visualize and analyze multiple runs, time-dependent
data, and forecast errors. A key contribution of this work is the comparative
visualization technique that allows users to analyze possible weather trends
and patterns. We illustrate the usage of our approach with a case study
designed and validated in conjunction with domain experts.
Visualization Framework for Inter-Media Comparison using Image Flows
Authors:
Masahiko Itoh, Masashi Toyoda, Cai-Zhi Zhu, Shinichi Satoh, Masaru Kitsuregawa
Abstract:
To understand recent societal behavior, it is important to compare how
multiple media react to real world events and how each medium reacts to other
media. This paper proposes a framework for inter-media comparison through
visualizing images extracted from different types of media. We extract blog
image clusters from our six-year blog archive and search for similar TV shots
in each cluster from a broadcast news video archive by using image
similarities. We then visualize such flows of images on a timeline to explore
visually changes in activities and interests of people and differences and/or
similarities between media such as image clusters that become hot topics on
only blogs or that become popular on blogs earlier than on TV.
Visualization of Passenger Flows on Metro
Authors:
Masahiko Itoh, Daisaku Yokoyama, Masashi Toyoda, Yoshimitsu Tomita, Satoshi Kawamura, Masaru Kitsuregawa
Abstract:
Visualization is one of the most important techniques for examining
influences of various kinds of phenomenon such as natural disasters, public
gatherings, or accidents on changes in behavior of Metro system passengers.
In this paper, we visualize the propagation of the effect of troubles and
changes in transportation flows in a wide area using data on the Tokyo Metro
extracted from a smart card system from March 2011 to April 2013. Our system
enables us to not only explore changes in passengers' actions after accidents
or disasters but also to discover unusual and unexpected phenomena and
explore their details and reasons.
Visualization to Facilitate Structured Exploration of Published Findings in Rat Brain Connectivity
Authors:
Hua Guo, Steven R.Gomez, Mark J. Schnitzer, David H. Laidlaw
Abstract:
We present the design, use cases, and user feedback for an online visual
analysis tool for integrating brain-network findings. We report needs and
challenges in studying brain connectivity as identified in close
collaboration with domain experts and describe how they inform the design of
a web-based tool for visualizing, annotating, and analyzing findings on
rat-brain connectivity. By representing both the hierarchical and
connectivity structure of the brain, the tool augments the original
information space through basic inference and facilitates structured
exploration of publications on brain connectivity. We finally report use
cases and open challenges in this domain as derived from user feedback by
neuroscientists. We found anecdotal support that neuroscientists using a
multi-scale visual analysis tool can find useful connection information more
quickly or discover more information than by using traditional searches like
PubMed queries.
Visualizing Dense Dynamic Networks with Matrix Cubes
Authors:
Benjamin Bach, Emmanuel Pietriega, Jean-Daniel Fekete
Abstract:
Visualizing static networks is already difficult, but exploring dynamic
networks is even more challenging due to the complexity of the tasks
involved; one visual encoding will hardly fit all tasks effectively, hence
multiple complementary views are needed. We introduce the Matrix Cube, a
visualization and navigation model for dynamic networks that results from
stacking adjacency matrices, one for each time step in the network. It builds
on our familiarity with cubes in the physical world and offers intuitive ways
to look at, manipulate and decompose them. We describe a set of operations to
decompose the Matrix Cube and interact with the resulting views.
Visualizing Geothermal Simulation Data with Uncertainty
Authors:
Sebastian Freitag, Bernd Hentschel, Torsten Kuhlen, Jan Niederau, Christian Vogt, Anozie Ebigbo, Gabriele Marquart
Abstract:
Simulations of geothermal reservoirs inherently contain uncertainty due to
the fact that the underlying physical models are created from sparse data.
Moreover, this uncertainty often cannot be completely expressed by simple key
measures (e.g., mean and standard deviation), as the distribution of possible
values is often not unimodal. Nevertheless, existing visualizations of these
simulation data often completely neglect displaying the uncertainty, or are
limited to a mean/variance representation. We present an approach to
visualize geothermal simulation data that deals with both cases: scalar
uncertainties as well as general ensembles of data sets. Users can
interactively define two-dimensional transfer functions to visualize data and
uncertainty values directly, or browse a 2D scatter plot representation to
explore different possibilities in an ensemble.
Visualizing Hidden Themes of Trajectories with Semantic Transformation
Authors:
Ding Chu, David A. Sheets, Ye Zhao, Yingyu Wu, Maogong Zheng, George Chen, Jing Yang
Abstract:
A new methodology, semantic transformation of taxi trajectory, is developed
to discover and analyze the hidden knowledge of massive taxi trajectories.
This approach creatively transforms the geographic coordinates (i.e. latitude
and longitude) to textual remarks. Consequently, each taxi trajectory is
studied as a document consisting of semantic information, such as the taxi
traversed street names, which enables semantic analysis of massive taxi data
sets as document corpora. Hidden themes, namely taxi topics, are identified
through topic modeling techniques. The taxi topics reflect urban mobility
patterns and trends, which are displayed and analyzed through a visual
analytics system. The system integrates interactive visualization tools such
as taxi topic maps, topic routes, street clouds, and parallel coordinates to
visualize the probability-based topical information. Urban planners,
administration, travelers, and drivers can conduct various knowledge
discovery tasks with direct semantic and visual assists. The effectiveness of
this approach is illustrated by case studies using a large taxi trajectory
data set acquired from 21,360 taxis in a city.
Visualizing Locations of Interest in 2D GPS Movement Data
Authors:
Shrey Gupta, Michael J. McGuffin, Thomas Kapler
Abstract:
We describe an interactive visualization of 2-dimensional movement data, such
as GPS data captured by smartphones. The raw data is analyzed to identify
places of interest (such as buildings visited) and to also identify meetings
between people (i.e., where two or more smartphones coincided in space and
time). The data is visualized using two coordinated views: a 2D geographic
map, and a Gantt chart. Design issues related to coordination of views and
visual representation of meetings are discussed.
Visualizing Sentiment Divergence Dynamics in Social Media Through SocialHelix
Authors:
Lu Lu, Nan Cao, Zhen Wen, Fei Wang, Yu-Ru Lin, Huamin Qu
Abstract:
Social media allow people to express and propagate different opinions, in
which people's sentiments to a subject often diverge when their opinions
conflict. An intuitive visualization that allows for unfolding the process of
sentiment divergence from the rich and massive social media data will have
far-reaching impact in various domains including social, political and
economic. In this poster, we propose a visualization system, SocialHelix, to
achieve this goal. SocialHelix is a novel visual design which enables users
to detect and trace occurring in social media, and to understand when and why
conflicts occurred and how they evolved among different social groups. We
demonstrate the effectiveness and usefulness of SocialHelix by conducting
in-depth case studies based on Twitter data regarding to the national
political debates.
Volume Rendering with Advanced GPU Scheduling Strategies
Authors:
Philip Voglreiter, Markus Steinberger, Rostislav Khlebnikov, Bernhard Kainz, Dieter Schmalstieg
Abstract:
Modern GPUs are powerful enough to enable interactive display of high-quality
volume data even despite the fact that many volume rendering methods do not
present a natural fit for current GPU hardware. However, there still is a
vast amount of computational power that remains unused due to the inefficient
use of the available hardware. In this work, we demonstrate how advanced
scheduling methods can be employed to implement volume rendering algorithms
in a way that better utilizes the GPU by example of three different
state-of-the-art volume rendering techniques.
Whale Sharks, Boolean Set Operations, and Direct Manipulation
Authors:
Ramik Sadana, Alistair Dove, John Stasko
Abstract:
Whale sharks are the largest form of fish and are on the list of vulnerable
species. We have developed a visualization technique to help marine
biologists explore blood samples taken from whale sharks and the different
bio-chemical compounds the samples contain. The visualization technique
models each sample as a set that may or may not contain the individual
compounds. Interactions on the samples show commonalities and differences as
well as patterns in compound presence. Biologists can use the visualization
to compare samples across days or weeks, find anomalies and trends resulting
from diet, and thus gain better insights into the health of the fish.