This Brilliant Visualization Could Build A Better Reddit

Web cluster

How Military Counterinsurgency Software Is Being Adapted To Tackle Gang Violence in Mainland USA

Cultural factors are more important than censorship in shaping Chinese surfing habits?

Web cluster

A Map Of Every College Mascot

Mascot Map

I’m not a college sports guy, but I do appreciate a good topic map that breaks down the relationships. Seems that I see “Bulldogs” everywhere I go and now I know why.

[A map of] mascots for more than 1,100 schools from Division I to Division III. Unsurprisingly, a good deal of real estate is dedicated to animals. It’s there you’ll find the Wolverines (in the “weasel family” subcategory of the “mammals” section), along with big names like the Minnesota Golden Gophers and some lesser-known ones like the Rhodes College Lynx. Bulldogs and Tigers are both significant hubs, though the Eagle looks to be the most popular mascot of all, with 60 schools claiming it as their avatar.

Network Theory Approach Reveals Altitude Sickness to be Two Different Diseases

Symptom Clusters

Being based in Boulder, altitude sickness is something we deal with when friends and relatives come to visit. Usually it’s not acute, thankfully, but even minor altitude sickness is no fun. Here is some new information coming out about the disease thanks to Graph Theory!

>Acute mountain sickness–or altitude sickness–is that common feeling of headache, dizziness, sleep disturbance and so on, that many people experience when travelling from lowland to high altitude.

>Kenneth Baillie at the University of Edinburgh […] used network theory to analyse the correlations between symptoms from a large number of people. The clusters in this network correspond to different diseases they say. The result could have an important impact on the way these diseases are treated and for the outcomes for victims.

Transformation of social networks in the late pre-Hispanic US Southwest

This is a paper on PNAS (behind a paywall) which I haven’t read yet.  So why would I post it?

Because the paper, titled “Transformation of social networks in the late pre-Hispanic US Southwest”, is about using “social network analysis with geographic information systems approaches to reconstruct network dynamics over 250” years to examine the demographic changes in the pre-Hispanic US Southwest.

That’s right, this is an anthropology/archaeology paper that is using social network analysis inferred though obsidian and ceramic artifacts to examine multigenerational changes in the population of the American Southwest.

This looks to be an excellent example of applying network theory to a problem outside of the usual areas of practice and finding some neat results.  For example, the abstract says “Our research shows how a large network in the southern Southwest grew and then collapsed, whereas networks became more fragmented in the northern Southwest but persisted.”

Links from Sears’ Talk

Sears just sent links to his slide deck and paper that he presented last week. Enjoy!


In many online social systems, interactions between individuals are observable but their friendships are hidden. Using behavioral data from 18 billion interactions among 17 million individuals and an anonymous online survey, we develop highly accurate predictors that automatically extract latent friendships between individuals with over 95% accuracy. Using these predictors we infer the hidden social network of individuals in the online game Halo: Reach and find that its structure is recognizably different from the popular online social network, Facebook. These results clarify the nature of friendship in online social environments and suggest new opportunities and new privacy concerns for friendship-aware applications that do not require the disclosure of private friendship information.

Decision Tree


graph-tool If you’re into the Python, you’ll wan to check this out.

graph-tool is an efficient python module for manipulation and statistical analysis of graphs (a.k.a. networks). Contrary to most other python modules with similar functionality, the core data structures and algorithms are implemented in C++, making extensive use of template metaprogramming, based heavily on the Boost Graph Library. This confers a level of performance which is comparable (both in memory usage and computation time) to that of a pure C/C++ library.

Marriage Problem and Variants

Solution to a Marriage Problem

This article is a nice overview of Marriage Problems on Bipartite Graphs, the algorithms for solving these problems, and variations you may encounter.

We discussed several meetings ago how the latest Nobel Prize in economics was awarded for the application of this class of problems in optimizing various situations.