Collaborative Graph Visualization System OnGraX Bjrn Zimmer and - - PowerPoint PPT Presentation

Displaying User Behavior in the Collaborative Graph Visualization System OnGraX

Björn Zimmer and Andreas Kerren

Department of Computer Science, ISOVIS Group, Linnaeus University, Sweden {bjorn.zimmer, andreas.kerren}@lnu.se

Content

• Collaborative Scenarios
• Motivation & Challenges
• OnGraX
• Video
• User Experiment
• Conclusion

2

3

Collaborative Scenarios

• P. Isenberg et al. “Collaborative Visualization: Definition,

Challenges, and Research Agenda,” Information Visualization, vol. 10, no. 4, pp. 310–326, 2011.

Motivation & Challenges 1/2

• Huge and growing amounts of data in social

sciences, software visualization, biology, …

• Experts in different locations want to explore

and analyze networks together

– Collaborative tools should be easy to use on the fly spontaneously and without setup overhead

• Collaborators should be able to establish and

keep a common ground1

1.

• J. Heer and M. Agrawala, “Design Considerations for Collaborative

Visual Analytics” Information Visualization, vol. 7, no. 1, pp. 49–62, Feb. 2008.

4

Motivation & Challenges 2/2

• Common Ground:

– Collaboration & Awareness:

• Who worked on the data previously?
• Where did the user look at?
• What changes were performed on the data?
• Was it an expert or a novice user?
• How to communicate thoughts/findings to subsequent users?

– Pointing & Reference:

• Participants of an analysis session can follow

activities of each other

5

OnGraX

• Collaborative graph visualization

in a web browser

• All actions are tracked
• Show current views and

mouse positions of users

• Annotations & chat links
• Revisit old graph states
• Visualization of user data

– Logged viewports – Changes to the graph

6

User Experiment

7 Node Changes Node Views

0.5 1 1.5 2 2.5 3 Glyphs Color Heat Map Mean Error Rate

25-30 min 15

Conclusion

• System for collaborative graph exploration

in a browser

– Seamlessly join and leave sessions - no additional applications needed

• Synchronous and asynchronous interaction

– Follow activities of other users and point to specific regions of the visualization – Visualize viewed regions and applied graph changes with heat maps

8

Thank You!

Any Questions?

9

Contact: bjorn.zimmer@lnu.se