1. Information Visualization Using 3D Interactive Animation Meng Tang 05/17/2001 George G. Robertson, Stuart K. Card, and Jock D. Mackinlay

2. Motivations  Goal  Lower the cost of finding information and accessing it  Strategies  Making the user’s immediate workspace larger  Enabling user interaction with multiple agents  Increasing the real-time interaction  Using visual abstraction to speed information assimilation and retrieval.

3. Technology Advances  3D graphics hardware  3D transformations, hidden-surface removal, double- buffered animation, antialiasing, lighting and surface models.  3D graphics libraries  OpenGL, PEX

4. Information Visualizer

5. Information Access vs. Document Retrieval  Document retrieval  Interest --> documents  Recall & Precision  Document retrieval is part of information use  Information is used to produce more information  New information is usually at a higher level of organization relative to some purpose

6. Information Workspaces  A virtual environment for finding information and accessing it.  Not just with the retrieval of information from a distant source, but also with the accessing of that information once it is retrieved and in use  Rooms System  Extend the desktop to multiple workspaces.  User can switch among multiple workspaces.

7. Information Workspaces

8.  Improving rooms system  Objective:  Decrease the costs for performing information-intensive tasks, or, alternatively, to increase the scope of information that can be utilized for the same cost.  Method:  Large Workspaces -- Make the immediate workspace virtually larger  Agents – Delegate part of the workload to semiautonomous agents  Real-Time Interaction – Maximize the interaction rates  Visual abstractions – Speed assimilation and pattern detection

9. Information Workspaces System GoalTechniques Large workspace to reduce access cost More screen space -> Rooms Denser screen space -> Animation, 3D Offload work to agents Search -> search agents Organizing -> clustering agents Interacting -> Interactive Objects Maximize real-time interaction rates Rapid interaction, Tune to human constants -> Cognitive Coprocessor schedule and Governor Visual abstraction to speed pattern detection Information Visualizations: Hierarchical structure -> Cone Tree Linear structure -> Perspective Wall Continuous data -> Data Sculpture Spatial data -> Office Floor Plan

10. UI Architecture  Several Problems  Multiple Agent Problem : How can system manage the interaction of multiple asynchronous agents.  Animation Problem : How can system provide smooth interactive animation  Interaction Problem : How can 3D widgets be designed and coupled to appropriate application behavior.  Viewpoint Movement Problem : How can the user changed the point of view rapidly and simply  Object Movement Problem : How can objects be easily moved about in a 3D space  Small Screen Space Problem : How can the dynamic properties of the system be utilized to provide the user with an adequately large work space.

11. UI Architecture Problem Solution Multiple Agent ProblemCognitive coprocessor scheduler Animation ProblemCognitive coprocessor scheduler and Governor Interaction ProblemInteractive objects Viewpoint Movement ProblemPoint of interest logarithmic flier Object Movement ProblemObject of Interest logarithmic manipulator Small Screen Problem3D/Rooms and 3D visualization

12. Cognitive Coprocessor  What is Cognitive Coprocessor  An animation loop and a scheduler for agents  An impedance matcher between the cognitive and perceptual information processing requirements of the user and the properties of these agents  3 sorts of time constants  Perceptual processing time constant (0.1sec)  Immediate response time constant (1sec)  Unit task time constant (5~30sec)

13. Cognitive Coprocessor  Perceptual processing time constant  Governor: reduce the quality to keep the frame rate.  Immediate response time constant  Agents provide status feedback at intervals no longer than this time constant  Immediate response animation  Unit task time constant  Time to complete a task  User can start the next request as soon as sufficient information has developed from the last request or even in parallel with it

14. Cognitive Coprocessor

15. Interactive Objects  Interactive objects  Basic building block in the Information Visualizer  Generalization of Rooms Buttons  2D/3D appearance  Allow mouse-based input (press, rubout, check, flick)

16. 3D Navigation and Manipulation  Overview  Doors  Walking metaphor  Point of interest logarithmic flight  Object of interest logarithmic manipulation

17. Information Visualization  Goal:  Attempts to display structural relationships and context that would be more difficult to detect by individual retrieval requests.  5 sorts of data organizations  Hierarchical: Cone Tree  Linear: Perspective Wall  Spatial: Structural browser  Continuous Data: Data sculpture  Unstructured: Information grid

22. Cone Tree

23.  Properties of Cone Tree  Each layer has cones of the same height  Cone base diameters for each level are reduced in a progression so that the bottom layer fits in the width of the room.  Body of each cone is shaded transparently  Text is shown only for the selected path  Manipulation of Cone Tree  Select nodes

24. Cone Tree  Advantages  the use of interactive animation to move some of the cognitive load to the human perceptual system  Using the depth to fill the screen with more information  Aspect ratio of cone tree is fixed  Fisheye view (brighter, closer, larger)  Disadvantages  Complex user interactions to access some of the information  fixing the aspect ratio limits the number of levels of the cone tree

26. Cone Tree

27. Perspective Wall  Obstacles to a visualization of linear information structures  The large amount of information that must be displayed  The difficulty of accommodating the extreme aspect ratio of the linear structure on the screen.  Disadvantages of Overview-Working View structure  Important contextual information, such as the neighborhood of the viewing region is just as small as unimportant details  Increase the space of overview leads to reduce the space of working view

29. Perspective Wall  Advantages  Integrate detailed and contextual views  Smooth transitions of views  Retain any 2D task-specific features  No special large- and small-scale versions of items must be designed  Ratio of detail an context is adjustable

31. Tracing Data Paths in 3D Graphics

32.  Number of errors in detecting paths in tree structures is substantially reduced if a 3D display method is used.  A tree structure is not necessarily a good candidate for 3D visualization