research V3 short - Fakultät für Informatik, TU Wien · Mor Nada Natalie Nikolaus Otto Panagiotis Paolo Paolo Patrick Peter Peter Peter Rene Rita Robert Robert Robert Roberto Ruth

Research Visual Analytics – VisualizationResearch WorkSilvia Miksch

Silvia Miksch: Short History

Linz

Backbone of Research Project Ecosystem

Visual Computing

Visual ComputingEXPAND

Doctoral College

www.cvast.tuwien.ac.atContent

Part 1:

Motivation & Contextualization

Visual Analytics of Time-Oriented Data

Challenges & Opportunities

Part 2:How Does Research Work

Research &TeachingScientific Writing Scientific Reviewing

´ Funding

Conclusion

Content

Part 1:






´ Funding

Conclusion

Visualization Success Story

Mystery: What is causing a cholera epidemic in London in 1854?

[Tufte, 1997]adapted from [Hearst , 2004]

London 18549

Which Information to Tackle ...

Patient Data

++

** *

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

GuidelinesData

Users

Tasks

... Change over Time

time

Patient Data

+

+

*

**

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

Guidelines

Patient Data

+

+

*

**

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

Guidelines

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

to do1. ..........................2. .......................... 3. ................

Guidelines

Patient Data

+

+

*

**

Data & Information Big Data

time

variablesOn the one hand, a huge amount of highly structured data and information is available in working situations and the daily life, ...

On the other hand, different kinds of data and information analysis methods were developed to gain more insights (information and knowledge gains).

time-oriented, multivariate, irregular sampled, having different temporal granularities, qualitative, quantitative, etc. structured and unstructured enriched by meta data

Motivation: Main Problems

Data Unmanageable – Information Overload

Missing Integration ofVarious (Heterogeneous) Information SourcesVarious Interdisciplinary Methods

Missing Involvement ofUsers and their Tasks

1 599 6932 525 6933 541 6624 542 6115 527 5796 505 5297 469 5538 409 5589 321 531

10 318 60611 321 69312 243 69313 250 66014 253 57915 246 52716 230 48917 196 51018 192 49719 134 50820 94 49321 25 42322 87 46723 128 48224 183 47325 163 568

26 541 55827 542 53128 527 60629 505 69330 469 69331 409 66032 321 57933 318 52734 321 48935 243 51036 250 49737 253 50838 246 49339 230 42340 196 46741 192 48242 134 47343 94 56844 541 55845 542 53146 527 60647 505 69348 469 69349 409 66050 321 579

51 318 52752 321 48953 243 51054 250 49755 253 50856 246 49357 230 42358 196 46759 192 48260 134 47361 94 56862 541 55863 542 53164 527 60665 505 69366 469 69367 409 66068 321 57969 318 52770 321 48971 243 51072 250 49773 253 50874 246 49375 230 423

76 196 46777 192 48278 134 47379 94 56880 318 55881 321 53182 243 60683 250 69384 253 69385 246 66086 230 57987 196 52788 318 48989 321 51090 243 49791 250 50892 253 49393 246 42394 230 46795 196 48296 318 47397 321 56898 243 48299 250 473

100 253 568

101 246 606102 230 693103 196 693104 230 660105 196 579106 318 527107 321 489108 243 606109 250 693110 253 693111 246 660112 230 579113 230 527114 196 489115 318 510116 321 497117 243 508118 250 493119 253 423120 246 467

Visualization for Problem Solving

Screen Resolution: 1024 * 768 = 786.432Yearly Measurements of Water Level in Low.Austria:1 5.256.000Number of Cellular Phones in Austria (2005):2 8.160.000Transmitted Emails Every Hours (World-Wide):3 35.388.000

Whole Data often not Presentable1. Applying Analytical Methods

(Data Reduction)2. Visualization of Most Important Data

and InformationAnalytical Methods

Statistics, Machine Learning & Data Mining

Analytical Methods

1 ... Amt der NÖ Landesregierung, Abt. WA5 - Hydrologie, http://www.noel.gv.at/SERVICE/WA/WA5/htm/wnd.htm2 ... CIA Factbook, https://www.cia.gov/cia/publications/factbook/3 ... How Much Information?, UC Berkeley, http://www2.sims.berkeley.edu/research/projects/how-much-info-2003/

today: peta (1015)tomorrow: exa (1018) &

zeta (1021)

Interactions

Past

Only passive ObservationsRepresentation not Changeable“one fits all”

Today

Active Examination with VisualizationsDynamically Adaptable and Modifiable→ Different Users, Tasks, and Aims

Visual Analytics

James Thomas & Kristin A. Cook NVAC (National Visualization and Analytics Center), Seattle, USA

“Visual Analytics is the science of analytical reasoning facilitated by interactive visual interfaces”

[Thomas & Cook 2005] [Keim, et al. 2010]

Visual Analytics – Process[Keim, et al., 2008]

Knowledge Generation Model for VA[Sacha, et al., 2014]

Knowledge Generation Model for VA[Sacha, et al., 2014]

Time has a Complex Structure


characterizingtime &

time-oriented data

visualizingtime-oriented

datainteracting

with timeanalyzing

time-oriented data

User-Centered Design

InteractiveVisual Analytics

Methods

data

goals/tasks user/audienceappropriateness

Content

Part 1:






´ Funding

Conclusion

CareCruiser Project

Interactive Exploration of Effects of Therapeutic Actions on a Patient’s Condition

Wolfgang Aigner, Theresia Gschwandtner, Katharina Kaiser, Silvia Miksch, Andreas Seyfang

[Gschwandtner, et al., 2010-2011]CareCruiser

[Gschwandtner, et al., 2010, 2011]

UsersMedical Experts & Physicians

DataPatient Data and Treatment Plans Data: multivariate, abstractTime: linear, instant

TaskExploring the Effects of Clinical Actions on a Patient’s Condition

Communicating Different Aspects with Multiple Views




Visualizing Temporal Aspects

Treatment Plans and Patient DataPatient Parameter Chart

Visualizing temporal aspects

Treatment Plans and Patient DataPatient Parameter Chart: Possible Value Range

maximum value

mimimum value

Visualizing temporal aspects

Treatment Plans and Patient DataPatient Parameter Chart: Intended Value Range

intended value range

Aligning Treatment Plans or Clinical Actions

38

Interacting with Time[Gschwandtner, et al. : CareCruiser]

CareCruiser Video

Interacting with Time[Gschwandtner, et al. : CareCruiser]

Lessons Learned

Visual Analytics: patient data in combination with applied treatment plans and clinical actions

Step-wise interactive exploration of effects New insights Generation of hypotheses

Evaluation (different points of view): Collaboration with medical expert guided designHeuristic Usability evaluation technical point of viewCase study domain point of view

Modeling Hypotheses with Visual Analytics Methods to Analyze the Past and Forecast the Future

Wolfgang Aigner, Markus Bögl, Tim Lammarsch, Silvia Miksch, Alexander Rind

Peter Filzmoser

[Lammarsch, et al., 2011, 2013], [Bögl, et al., 2011, 2013]

TiMoVA Visual Analytics for Model Selection in Time Series Analysis

Users Experts in time series analysis

DataData: univariate, abstractTime: instant

TasksTime series transformation and model selection

Example: Statistical SW Tool Gretl Preview-Video[Bögl, et al., 2011, 2013]

Usage Scenario Model SelectionTime Series Line Plot

Autocorrelation Function (ACF) Residual Analysis

[Bögl, et al., 2011, 2013]

TiMoVA – VA Prototype

Definition of

[Shumway and Stoffer, 2011]

[Bögl, et al., 2011, 2013]

TiMoVA – VA Prototype[Bögl, et al., 2011, 2013]

Usage Scenario Model SelectionTime Series Line Plot

Autocorrelation Function (ACF) Residual Analysis

[Bögl, et al., 2011, 2013]

Lessons Learned

EvaluationApplied usage scenarios based on the requirementsFormative evaluation during design and implementation phaseDemonstration session internally and with two external experts

Supports and guides domain experts by

Model order selectioninside the relevant plotImmediate visual feedbackof the model residualsVisualization of the model transitionsShort visual feedback cycles

Visual Analytics of Time‐Oriented Data


Infrastructure for Reusable Components

TimeBench: A Data Model and Software Library

Infrastructure to Facilitate Evaluation

EvalBench: A Software Library for Visualization Evaluation

Science of InteractionRole and Value of Interactivity

Information Discovery Process

Usability & Utility

Insights Studies

Knowledge Respositories

Supported by the Austrian BMWFJ via CVAST, a Laura Bassi Centre of Excellence (#822746), and by the Austrian Science Fund (FWF) viaHypoVis (#P22883)

Multiple Granularities Different Time Primitives Temporal Indeterminacy

[Rind, et al., 2013]

Reduces implementation effort for evaluation featuresConsistent and reproducible execution of study protocols

Integrates well with existing visualization prototypesFree and open source software (@GitHub)

Supports:Controlled Experiments

Interaction LoggingLaboratory Questionnaires

Heuristic EvaluationsInsight Diaries

[Aigner, et al., 2013] How to use?[Aigner, et al., 2013]

How to use?

Define task lists for sessions

[Aigner, et al., 2013]

How to use?

Define task lists for sessionsImplement EvaluationDelegate interface


How to use?

Define task lists for sessionsImplement EvaluationDelegate interface


Content

Part 1:






´ Funding

Conclusion

Challenges

(Time-Oriented) DataScale and Complexity

Heterogeneous Data(Meta Data, Semantics, Multiple Sources, etc.)

Data Quality & Uncertainty

Data Provenance

DesignGuidance on how to Design and Develop Visual Analytics Systems

UserMeeting Users’ Needs

High Degree of Interactivity(Temporal Dimensions)

Evaluation (Qualitative & Quantitative)

TechnologyProvide Reusable Infrastructure

[VisMaster Challenges 2010]

Application Challenges in Visual Analytics

Challenges

(Time-Oriented) DataScale and Complexity

Heterogeneous Data(Meta Data, Semantics, Multiple Sources, etc.)

Data Quality & Uncertainty

Data Provenance

DesignGuidance on how to Design and Develop Visual Analytics Systems

UserMeeting Users’ Needs

High Degree of Interactivity(Temporal Dimensions)

Evaluation (Qualitative & Quantitative)

TechnologyProvide Reusable Infrastructure

[VisMaster Challenges 2010]

Application Challenges in Visual Analytics

Challenges: Visual Analytics – Process[Keim, et al., 2008]

ConclusionVisual Analytics –

Detect the Expected and Discover the Unexpected

InteractiveVisual Analytics

Methods

data

goals/tasks user/audienceappropriateness

Thanks to

Alan Albert Alessio Alexander Alexander Alime Amin Andreas Andreas Annette Arghad Barbara Barbara Ben Bilal Brain Burcu Carlo Catherine

Christian Christian Christian Claudio Daniel David Dorna EdeltraudEduard Elisabeth Elpida Elske Eva Fabian Felix Florian Florian Frank

Franz Gennady Georg Georg Gerhard Gerhilde Guiseppe Hanna Heidrun Helga Helwig Ingrid Jarke Jim Jimmy Johannes Jörn Jürgen Kai Karl

Katharina Klaus Krist Luca Lukas Manfred Mar Margit Maria Markus Markus Martin Martin Matt Michael Michael Michael Mikko Monika Monika

Mor Nada Natalie Nikolaus Otto Panagiotis Paolo Paolo Patrick PeterPeter Peter Rene Rita Robert Robert Robert Roberto Ruth Sabine SalvoSamson Silvana Simone Sophie Stefan Stefan Stephan Susanne Sylvia

Taowei David Theresia Thomas Tim Tom Werner Wolfgang Yuval

... and many students and co-workerswww.timeviz.net

Wolfgang Aigner • Silvia MikschHeidrun Schumann • Christian Tominski

Visualization of Time-Oriented Datawith a foreword by Ben Shneiderman

Springer1st Edition., 2011, XVIII, 286 p. 221 illus., 198 in color.Hardcover, ISBN 978-0-85729-078-6Due: June 2011

Table of ContentsIntroduction • Historical Background • Time & Time-Oriented Data • Visualization Aspects • Interaction Support • Analytical Support • Survey of Visualization Techniques • Conclusion

www.timeviz.net Horizon Graph[Reijner, 2005]

HOW DOES RESEARCH WORKPart 2

Content

Part 1:






´ Funding

Conclusion

7

WARUM habe ich

WAS gemacht,

WIE habe ich es gemacht und

mit WELCHEM ERGEBNIS.

[Hevner et al. 2004]

[Hevner et al. 2004] [Hevner et al. 2004]

Research & Teaching

Bachelor Course

Master Courses

Supervision of Bachelor Students

Supervision of Master Students

Supervision of PhD Students

Publications

Conferences

Journals

Contributions in &Books

[Simon L. Peyton Jones, 2004 presentation]

>> Title, Affiliation <<

>> References <<

Different in each community

Different ...

Different for

JournalsConferenes

Timely and Complex Process

Steps

..... Example EuroVis 2012 & VisWeek 2012

Peer Reviewing Process

7

PC

IPC

Roles

20.4.2 188.917 – VU Informationsdesign und Visualisierung 8

8

Primary and Secondary Reviewers

8

Review Forms

8

Review Forms

8

Review Forms

8

Review Forms

8

Review Forms

8

Be TimelyProtect IdeasAvoid Conflict of InterestBe SpecificBe HelpfulBe Tactful(In Summary)

Ethics Guidelines

8

http://vgtc.org/wpmu/techcom/conferences/ethics-guidelines/#In%20Summaryhttp://www.uib.no/eurovis2011/reviewing_guidelines.php

Funding

Research needs funding ...

... Basic Research

... Applied/Coooerative Reserch

Backbone of Research Project Ecosystem

Visual Computing

Visual ComputingEXPAND

Doctoral College

www.cvast.tuwien.ac.at

Thanks to (Intern)ational Collaborations

Alan Albert Alessio Alexander Alexander Alime Amin Andreas Andreas Annette Arghad Barbara Barbara Ben Bilal Brain Burcu Carlo Catherine

Christian Christian Christian Claudio Daniel David Dorna EdeltraudEduard Elisabeth Elpida Elske Eva Fabian Felix Florian Florian Frank

Franz Gennady Georg Georg Gerhard Gerhilde Guiseppe Hanna Heidrun Helga Helwig Ingrid Jarke Jim Jimmy Johannes Jörn Jürgen Kai Karl

Katharina Klaus Krist Luca Lukas Manfred Mar Margit Maria Markus Markus Martin Martin Matt Michael Michael Michael Mikko Monika Monika

Mor Nada Natalie Nikolaus Otto Panagiotis Paolo Paolo Patrick PeterPeter Peter Rene Rita Robert Robert Robert Roberto Ruth Sabine SalvoSamson Silvana Simone Sophie Stefan Stefan Stephan Susanne Sylvia

Taowei David Theresia Thomas Tim Tom Werner Wolfgang Yuval

... and many students and co-workers

Content

Part 1:






´ Funding

Conclusion

TIPS & TRICKSPart 3

TIPS: How to Do Research

9

http://www.ifs.tuwien.ac.at/~silvia/research-tips

TIPS: How to Do Research TIPS: How to Do Research


siehe



My Favorite Book for years ...

1

Lyn Dupré:

BUGS in Writing:a Guide to Debugging Your Prose, Addison-Wesley, 1998

... and

1

William Strunk Jr., E. B. White

The Elements of StyleLongman, New York; 4th edition 1999

1

WARUM habe ich

WAS gemacht,

WIE habe ich es gemacht und

mit WELCHEM ERGEBNIS.

research V3 short - Fakultät für Informatik, TU Wien · Mor Nada Natalie Nikolaus Otto Panagiotis Paolo Paolo Patrick Peter Peter Peter Rene Rita Robert Robert Robert Roberto Ruth

Documents