INFORMATION VISUALIZATION
Visualization
The use of computer-supported, interactive, visual representations of data to amplify cognition.
The purpose of visualization is insight not pictures
Goals of insight are Decision making, Discovery and Explanation
Why Visualization?
A picture is worth ten thousand words Amplify our cognition ability
Cognition: the acquisition or use of knowledge Specific goals:
Communicating ideas Create and discover ideas Use visual perception to solve problems
To get a ‘Ah HA’ response from the viewer
Origin
Data graphics-1786-Playfair(Use lines, areas visually)
Theory of Graphics-1967-Bertin(Plotting Data)
Theory of Data-1983-Tufte(maximising density of useful information)
Exploratory Data Analysis-use of pictures to give statistical insight to Data
Visualization amplifies cognition
Increases resources Reduces search Enhanced recognition of patterns Perceptual inference Perceptual monitoring Manipulable medium
Visualization Principles
Expressiveness: Encode all the facts in the result set. Encode only the facts in the result set.
Effectiveness: Depends on the capability of the perceiver. Encode the more important information more
effectively.
Visualization – Twin Subjects
Scientific Visualization Visualization of
physical data
Information Visualization Visualization of
abstract data
Ozone layer around earthAutomobile web site- visualizing links
Scientific Visualization – Information Visualization
Focus is on visualizing set of observations that are multi-variate
There is no underlying field – it is the data itself we want to visualize
The relationship between variables is not well understood
Focus is on visualizing an entity measured in a multi-dimensional space
Underlying field is recreated from the sampled data
Relationship between variables well understood
Scientific Visualization
Information Visualization
Information Visualization
“… is a process of transforming data and information that are not inherently spatial, into a
visual form allowing the user to observe and understand the information.”
(Source: Gershon and Eick, First Symposium on Information Visualization)
“… the use of computer-supported, interactive, visual representations of abstract data to amplify cognition.”
Card, Mackinlay, Shneiderman
Basic Visualization Model
Goal
Data
Data transfer
Insight(learning, knowledge extraction)
Method
Data
Visualization
Map: data → visual
~Map-1: visual → data insight
Data transfer
Insight
Visual transfer
(communication bandwidth)
Visual Mappings
Data
Visualization
Map: data → visual
Visual Mappings must be:• Computable (math)
visual = f(data)
• Comprehensible (invertible)data = f-1(visual)
• Creative!
Visualization Pipeline
Raw data(information)
Visualization(views)
Data tables
Visualstructures
Datatransformations
Visualmappings
Viewtransformations
task
User interaction
Visual Mapping: Step 1
1. Map: data items visual marks
Visual marks: Points Lines Areas Volumes Glyphs
Visual Mapping: Step 2
1. Map: data items visual marks2. Map: data attributes visual properties of
marksVisual properties of marks:
Position, x, y, z Size, length, area, volume Orientation, angle, slope Color, gray scale, texture Shape Animation, time, blink, motion
Information Types
Multi-dimensional: databases,… 1D: timelines,… 2D: maps,… 3D: volumes,… Hierarchies/Trees: directories,… Networks/Graphs: web, communications,… Document collections: digital libraries,…
1-D Data
Linear data: textual document, source code, etc.
User problems: count, find, replace, … Encoding: fonts, color, size, layout, scrolling,
selection capabilities, … Product example: text editor, browser, …
2-D Data
Planar or map data: geographical maps, floor plans, newspaper layouts, …
User problems: find adjacent items, search containment, find paths, filtering, details-on-demand, …
Encoding: size, color, layout, arrangement, multiple layers, …
Product example: CAD
3-D Data
Real-world objects: building, human body User problems: adjacency in 3-D,
inside/outside relationship, position, orientation
Encoding: overviews, landmarks, transparency, color, perspective, stereo display
Product example: CAD
Temporal Data
Time series data: medical records, project management, historical presentation
User problems: finding all events before, after or during some time period or moment.
Encoding: time lines
Multi-dimensional Data
Relational and statistical databases tuples. User problem: finding patterns, clusters,
correlations, gaps, outliers. Challenge:
Simultaneously display many dimensions of large subsets of data.
Create displays that best encode the data pattern for a particular task.
Rapidly select a subset of tuples or dimensions.
Tree Type Data
Exponential data: hierarchies, tree structures. User problems: find the structural properties
Height of the tree Number of children Find nodes with same attributes
Encoding:Node-link diagrams: allowing the encoding of linkage between entities. Treemap: child rectangles inside parent rectangles
Product example: windows explorer, internet traffic
Network Data
Graph data: multiple paths, cycles, lattices User problems:
Shortest path Topology problems
Encoding: Node-link diagram Matrix
Basic Visualization Tasks
Overview of a collection of data. Zoom in/on objects of interest. Filter out uninterested items. Details-on-demand: view details. Relate: View relationship. History: Undo, Redo, Refinement. Extract a subset of the data.
User Tasks
Easy stuff: Min, max, average, % These only involve 1 data item or value
Hard stuff: Patterns, trends, distributions, changes over time, outliers, exceptions, relationships, correlations, multi-way, combined min/max, tradeoffs, clusters, groups, comparisons, context, anomalies, data errors, Paths, …
Excel can do this
Visualization can do this!
Scientific Visualization Model
visualizemodeldata render
Visualization represented as pipeline:Read in dataBuild model of underlying entityConstruct a visualization in terms of geometryRender geometry as imageData are more spatial
Classification of InfoVis Techniques
Based on the type of information Visualization of Information Structure
Trees, Networks Visualization of Multivariate Data
1D, 2D, 3D, n-D, Temporal Visualization of Workspace
Windows, web pages, documents, etc
Classification of InfoVis Techniques
Based on how we interact with the data Overview: fisheye Zooming: e.g. Table Lens Interactive filtering: e.g. Magic Lens Brushing and linking: e.g. XGobi Details-on-demand: e.g. Spotfire
InfoVis Design Issues
Selection What data should we choose to visualize?
Representation How should data be represented? Colors? Locations?
…
Presentation Too much data, too little display space
InfoVis Design Issues (cont’d)
Scale and Dimensionality What if you have 93 variables to visualize?
Interaction and Exploration How user interacts with the data?
Hot Topics: network visualization, document visualization, security related problems, etc
InfoVis Applications
Complex Documents Biography, manuscript, data structure
Histories Patient histories, student records, etc
Classifications Table of contents, organization charts, etc
Networks Telecom connections and usages, highway, etc
InfoVis Example
Stephen Eick’s Seesoft
InfoVis Example
Hyperbolic Trees
2D Hyperbolic Tree 3D Hyperbolic Tree
Escher’s woodcut
InfoVis Example
Themescape (Cartia)
Interactive Graphics
Homefinder
Visualization Techniques
Location Probes-eg.Film finder(use location to view additional data)
View point Controls-eg.Information mural(Overview+detail)
Distortion-eg.Perspective Wall(Focus+Context)
Focus + Context
Data Types
Univariate Bivariate Trivariate
Univariate
Dot plot Bar chart (item vs. attribute) Histogram
Bivariate
Scatterplot
Trivariate
3D scatterplot, spin plot 2D plot + size (or color…)
Information Visualization Using 3D Interactive
Animation
Visual Abstractions
Hierarchical Structure -> Cone Tree Linear Structure -> Perspective Wall Continuous Data -> Data Sculpture Spatial Data -> Office Floor plan
Cone Tree
Cone Tree (Compaq Research)
Perspective Wall
Spatial data
Web Book and Web Forager. Card, Robertson, York. 1996.
Document Visualization
IE Vs IV
Examples 1-D
Traversing long lists in changeable sort orders. Viewing summary data about many ordered items, possibly to find important specific elements. Filtering out unwanted items.
2-D Semantic zooming
3-D Containment issues. Position (Up, Down, Inside, Outside) queries.
Multi-D Understand or get an overview of the whole or a part of the n-dimensional data. For example, finding patterns, relationships, clusters, gaps, and outliers of the data. Find a specific item in the data. For example, zooming, filtering and selecting a group or a single item from the data.
Temporal Viewing events or data in sequence and/or varying the order of those events or data. Viewing and creating historical overviews of events or data. Finding temporal inconsistencies and/or undesirable relationships in events or data.
Tree Obtain global relationships and structure from the entire hierarchy. Find the most recent common ancestor between two nodes. Find the path to a particular node from the root of the hierarchy. Find clusters, duplicates, relationships, and inheritance properties from the structure of the hierarchy. Discovering attributes (especially the size) of nodes or entire subtrees.
Network Change the layout of nodes of a graph so that the network is easier to comprehend. Relate: enable the user to find relationship among the nodes in a diagram. Find interesting paths in a graph.
Workspace Enable users to view and interact with computer screen layouts in a more efficient manner. Allow geographically dispersed users to collaborate and interact concurrently. Synthesize information, expertise, and results to create high quality solutions. Organize, interact, and search task related information efficiently. Allow rapid access and restructuring for task information.
Data Scalability
# of attributes (dimensionality)
# of items
Value range(e.g. bits/value)
Increase Data Density
Calculate data/pixel
“A pixel is a terrible thing to waste.”
(Tufte)
(Shneiderman)
Advantages
Reduce visual search time (e.g. by exploiting low level visual perception).
Provide a better understanding of a complex data set (e.g., by exploiting data landscape metaphors).
Reveal relations otherwise not being noticed (e.g., by exploiting the mind’s ability to see relationships in physical structures).
Enable to see a data set from several perspectives simultaneously.
Are effective sources of communication.
Visualization techniques can help individuals to abstract, customize, manipulate and understand the information being presented
Conclusion
Visualization helps Information presentation Information extraction
Good visual encoding should match the target data and user problems.
Studying the successful/unsuccessful visual encoding designs and techniques helps us to design and develop new encoding approaches.
IE Vs IV
Topics
Information Types: Multi-D 1D 2D 3D Hierarchies/Trees Networks/Graphs Document collections
Strategies: Design Principles Interaction strategies Navigation strategies Visual Overviews Multiple Views Empirical Evaluation Development Theory Tools
Philosophy: Optimization
Visualization = the best of bothImpressive computation + impressive cognition
Computer•Serial•Symbolic•Static•Deterministic•Exact •Binary, 0/1•Computation•Programmed •Follow instructions•Amoral
Human•Parallel •Visual •Dynamic •Non-deterministic •Fuzzy•Gestalt, whole, patterns •Understanding •Free will•Creative •Moral
Thank You