Visual Analytics Techniques that Enable Knowledge Discovery:
description
Transcript of Visual Analytics Techniques that Enable Knowledge Discovery:
![Page 1: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/1.jpg)
1
Visual Analytics Techniques that Enable Knowledge Discovery:
Detect the Expected and Discover the Unexpected
Jim J. ThomasDirector, National Visualization and Analytics Center
AAAS Fellow, Pacific Northwest National Laboratory Fellowhttp://[email protected]
ACM SIGKDD Workshop on Visual Analytics and Knowledge Discovery
VAKD '09 Paris, France
![Page 2: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/2.jpg)
Visual Analytics Techniques that Enable Knowledge Discovery
Introduction: what is and is not visual analytics?Landscape of visualization scienceDiscussion of selected existing systems and technologiesCommon characteristics enabling knowledge discoveryTop ten challenges
2
![Page 3: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/3.jpg)
Introduction:History of Graphics and Visualization• 70s to 80s
– CAD/CAM Manufacturing, cars, planes, and chips– 3D, education, animation, medicine, etc.
3
• 80s to 90s– Scientific visualization– Realism, entertainment
• 90s to 2000s– Information visualization– Web and Virtual environments
• 2000s to 2010s– Visual Analytics– Visual/audio analytic appliances
![Page 4: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/4.jpg)
Visual Analytic Collaborations
Detecting the Expected -- Discovering the UnexpectedTM
4
![Page 5: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/5.jpg)
Visual Analytics Definition
Visual analytics is the science of analytical reasoning facilitated by interactive visual interfaces.
People use visual analytics tools and techniques to Synthesize information and derive insight from massive,
dynamic, ambiguous, and often conflicting data Detect the expected and discover the unexpected Provide timely, defensible, and understandable assessments Communicate assessment effectively for action.
“The beginning of knowledge is the discovery of something we do not understand.” ~Frank Herbert (1920 - 1986) 5
![Page 6: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/6.jpg)
What is not visual analytics?
Large graph structure with no labelsHeat map with no labelsSearch and retrieval systemsChart with no interactionImage with no semantic interpretationStand alone image that does not tell a story
6
![Page 7: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/7.jpg)
The Landscape of Visualization Science
Publications from IEEE VisWeek, 2006, 2007, 2008using IN-SPIRE Visual Analytics Tool
Each dot is an published science article, full text
![Page 8: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/8.jpg)
Systems Considered:IN-SPIRE - http://in-spire.pnl.gov.
JIGSAW - John Stasko, Carsten Görg, and Zhicheng Liu, “Jigsaw: Supporting Investigative Analysis through Interactive Visualization,” Information Visualization, vol. 7, no. 2, pp. 118-132, Palgrave Magellan,
2008.WIREVIZ - Remco Chang, Mohammad Ghoniem, Robert Korsara, William Ribarsky, Jing Yang, Evan Suma, Carolina Ziemkiewicz, Daniel Keim, Agus Sudjianto, IEEE Visual Analytics Science and Technology
(VAST) 2007. GreenGrid - Pak Chung Wong, Kevin Schneider, Patrick Mackey, Harlan Foote, George Chin Jr., Ross Guttromson, Jim Thomas “A Novel Visualization Technique for Electric Power Grid Analytics,” IEEE Transactions on Visualization and Computer Graphics 15(3):410-423.
Scalable Reasoning System - Pike WA, JR Bruce, RL Baddeley, DM Best, L Franklin, RA May, II, DM Rice, RM Riensche, and K Younkin. (2008) "The Scalable Reasoning System: Lightweight Visualization for Distributed Analytics." In IEEE Symposium on Visual Analytics Science and Technology (VAST).
8
![Page 9: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/9.jpg)
Whole - Part RelationshipScale independent representations, whole and parts at same time at multiple levels of abstraction, often linked
9
![Page 10: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/10.jpg)
Whole - Part Relationship
10
![Page 11: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/11.jpg)
Relationship DiscoveryExplore high dimensional relationships, theme groupings, outlier detection, searching by proximity at multiple scales
11
![Page 12: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/12.jpg)
Relationship Discovery
12
Boolean
By Example
![Page 13: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/13.jpg)
Combined Exploratory and Confirmatory Analytics
Develop and refine hypothesisEvidence collection, management, and matching to hypothesisTailor views/displays for thematic/hypothesis focus of interestOften suggestive of predictions enabling proactive thinking
13
![Page 14: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/14.jpg)
Multiple Data Types
Supports multiple data types: structured/unstructured textImagery/video, cyberSystems of either data type or application specific
14
![Page 15: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/15.jpg)
Temporal Views and InteractionsMost analytics situations involve time, pace, velocityGroup segments of thoughts by timeCompare time segmentsOften combined with geospatial
15
![Page 16: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/16.jpg)
Reasoning Workspace
16
Workspace to construct logic and illustrate reasoningFlexible spatial view of reasoning: stories
Stu Card, PARC
![Page 17: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/17.jpg)
Grouping and Outlier DetectionForm groups of thought/dataLabels and annotationCompare groupingsFind small groups or outliers
17 17
![Page 18: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/18.jpg)
LabelingCritically important, Dynamic in scope, number labels, size, colorPositioningAlmost everything has labelsLabels tell semantic meaning
18
![Page 19: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/19.jpg)
Multiple Linked ViewsTemporal, geospatial, theme, cluster, list views with association linkages between views
19
![Page 20: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/20.jpg)
Multiple Linked ViewsTemporal, geospatial, theme, cluster, list views with association linkages between views
20
Heatmap View(Accounts to Keywords Relationship)
Strings and Beads(Relationships over Time)
Search by Example (Find Similar Accounts)
Keyword Network(Keyword Relationships)
![Page 21: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/21.jpg)
WireViz Video
21
![Page 22: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/22.jpg)
ReportingCapture display segments in graph modes for putting in reports, PPT etcCapture reasoning segments of analytic resultsCapture animations
22
![Page 23: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/23.jpg)
Engaging InteractionGreenGrid video
23
Alberta
North California
Southern
Northern
![Page 24: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/24.jpg)
GreenGrid Video
24
![Page 25: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/25.jpg)
Tested With Known Data and Solutions
25
![Page 26: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/26.jpg)
26
Top Ten Challenges Within Visual Analytics
Human Information Discourse for Discovery—new interaction paradigm based around cognitive aspects of critical thinkingNew visual paradigms that deal with scale, multi-type, dynamic streaming temporal data flowsData, Information and Knowledge RepresentationCollaborative Predictive/Proactive Visual AnalyticsVisual Analytic Method Capture and Reuse
![Page 27: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/27.jpg)
27
Top Ten Challenges Within Visual Analytics
Dissemination and CommunicationVisual Temporal AnalyticsValidation/verification with test datasets openly availableDelivering short-term products while keeping the long viewInteroperability interfaces and standards: multiple VAC suites of tools
27
![Page 28: Visual Analytics Techniques that Enable Knowledge Discovery:](https://reader034.fdocuments.net/reader034/viewer/2022052510/56815b88550346895dc98cf9/html5/thumbnails/28.jpg)
Conclusions
Visual Analytics is an opportunity worth consideringPractice of Interdisciplinary Science is requiredBroadly applies to many aspects of society For each of you:
28
The best is yet to come…