Much of my research at the Media Laboratory centers on online social interaction. Recently, I have been looking at IRC, trying to understand how interaction on different channels evolves over time. I propose a system that visualizes, in 3D, statistical information about a collection of IRC channels. As part of my research, I am building an active, kinetic, qualitative 2D visualization of an IRC channel. By working on the alternative 3D statistical representation for this class, I hope to derive a different, deeper understanding of the data I am using for my Masters research.

Some of the statistical data that I hope to incorporate into my 3D visualization include:

• Number of users over time
• Log on rate
• Log off rate
• Average post length
• Post rate
• Population growth rate
• Percentage of population that post

There are, of course, other measurements that will be used. Hopefully, by starting with the above listed statistics, other important statistics will be suggested by the visualization, which can then be incorporated. I also hope to incorporate data about individual users into the visualization.

I will complete the following work by the dates listed below:

4.14.2000

Gather initial statistical data
Format data as appropriate for import into VTK
Create initial test visualizations in VTK

4.28.2000

Finalize visualizations, deciding on angle, color, and mapping
Build interactive visualization in java if feasible
Begin analysis of quality of visualization, including ease of reading and answering question: "What does this visualization allow us to understand about the social interaction in IRC channels?"

5.8.2000

Finish web page for final presentation
Complete the analysis of the visualization
Finish final project report

Visualization: This will ideally be an interactive 3D visualization of the statistical data outlined in the summary. If it becomes clear that interactivity is unattainable, then there should at least be a time-lapse movie showing the data for some segment of time--hopefully one day?s worth of data.

Analysis: A complete analysis of the success of the visualization, answering such questions as: "Is the visualization easily readable?", "Does the interaction (or motion) of the visualization add to the understanding of the underlying patterns?", and "What does the visualization help us to understand about interaction patterns in chat systems?"

Please note that while the project proposed above does draw from the same data that I am using to build my Masters Thesis project, the visualization and analysis is a new and different direction of research in this area. I hope that this work will aid in my research into social interaction online.