Workshop on Statistical Inference, Computing and Visualization for Graphs
August 1 - 2, 2003
Stanford University

 

 

Home Page

Participants Logistics Program

 


Links to :  Contributed Talks ---  Software Demos --- Printable Version of Program


note: slides and papers are all in PDF unless noted otherwise

Day 1- Friday, August 1

8:30-8:45 Welcome
Applications and Challenges
8:45-9:15 Transactional Data Overview Daryl Pregibon
9:20-9:50 Social Networks Overview (slides) Mark Handcock
10:00

Break

10:30-11:00 Link Mining (slides.ppt) Lise Getoor
11:00-11:30 Bioinformatics Overview (slides) Susan Holmes
11:30-12:15 Group Discussion led by Robert Gentleman
12:15-1:30

Lunch

Software and Algorithms
1:30-2:00 Graph Drawing and Layout Overview (slides)

Stephen North

2:00-2:30 Graph Algorithms Overview (slides.ppt) Sridhar Rajagopalan 
2:30-3:00

Break

3:00-3:20 Graph Drawing and Analysis in R (slides) Vince Carey
3:20-3:40 GGobi (slides) Debby Swayne
3:45-4:15 Group Discussion led by Duncan Temple Lang
4:30-??? Demos:

Zoomgraph Eytan Adar and Josh Tyler (link to demos, downloads, paper)

Visualizing Augsburg Traffic Data with VanGogh  Rene Keller (paper, website)

Exploratory Data Analysis of Graphs in GGobi Deborah F. Swayne  (slides)

Day 2 - Saturday, August 2

Exploratory Data Analysis
8:30-8:45 Welcome again
8:45-9:15 Exploratory Data Analysis with Graphs (slides.ppt) Chris Volinsky
9:15-9:45 Exploratory Data Analysis-Bioinformatics  (slides) Robert Gentleman
9:45-10:30

Discussion and Break

Statistical Inference and Modeling with Graph Data - Research Talks - (click for abstract)
10:30 - 12PM

Lada Adamic - HP Labs  "A social network caught in the Web" (slides, paper)

Deepak Agarwal - AT&T Labs "Statistical Inference for Large Directed Graphs with Communities of Interest" (slides)

Mark S. Handcock - University of Washington "Degeneracy and Inference for Social Networks Models"

Denise Scholtens - Harvard School of Public Health "Graph theory and statistical inference considerations for protein-protein interaction and gene expression data" (slides)

2:00-1:30 Lunch
1:30 - 3:00 Cliff Behrens - Telcordia, Inc "Graphical Representations of Knowledge and Its Distribution" (slides.ppt)

Friedrich Leisch - Universität Wien  "Trellis-Style Displays with a Graph Layout in R" (slides)

Daniel A. McFarland - Stanford Universtiy "Dynamic Network Visualization: Methods for Meaning with Longitudinal Network Movies" (slides.ppt)

Scott White - University of California - Irvine "A Brief Guided Tour of the Java Universal Network/Graph Framework" (web page, slides)

3:00 Wrapup Discussion - Andreas Buja

Day 1, morning: Applications and challenges

Presentations on different application areas: What does the data look like? What approaches are being used, and what challenges does the data pose? These should be high level talks that introduce the field to the unfamiliar, and present the particular challenges faces with regard to using graphs for statistical analysis.  

Day 1, afternoon: Software and Algorithms

Presentations on existing software that can be used for graphs, the strengths and limitations and challenges for the graph drawing comminity which are specific to statistical analysis.  We will have representatives from the graph algorithms community as well as the statistical software community -- how can we bring these two together? 

Graph Software - What tools are in use now? What's good about them, and where do they fall short?

Statistical software for Graphs - where do statisticians interface with the technologies above?

Day 2, morning and early afternoon: Statistical Inference and EDA for Graphs

Discussions of the various types of analysis we are doing with graphs, with specific examples.  This section would give participants an opportunity to give a short talk on their specific research, with the hope that it ties in to previous sections. 

Day 2, late afternoon: Summing up

One topic for this final session is a conversation between users and software designers: what requests do users have for the software designers, and what questions would the designers like to ask of the prospective users?