Co-authorship within several fields
http://www.pnas.org/content/pnas/101/suppl_1/5200.full.pdf
Summary: Co-authorship networks are one of the standard examples of a real world network. The vertices of the graph are the authors themselves and edges connect two who have co-authored a paper together. As with any network dataset some natural questions arise such as how connected is the network? How does this compare across various academic fields? This paper examines some datasets withing co-authorship and gives analysis. They look at biomedical, physics and mathematics co-authorsip networks with the biomedical network being by far the largest, despite spanning a much shorter time period than the others. Interestingly, the mathematics and physics authors typically had very few co-authors but a select few would have upwards of hundreds or thousands. This phenomenon is known as heavy tailed distributions in networks and while the paper adresses this minimally it is a widely discussed concept within the field of networks.
How this realtes to class:
The class so far has layed the foundational work to discuss networks with ideas such as nodes, edges, and degree of nodes. This paper is one of many examples of how one can do valuable analysis of real networks with this simple but powerful set of tools. The paper also discusses the triadic closure property that we covered in class. They found that one in biomedical research is far less likely than in mathematics or physics to have co-authors that are also co-authors themselves. In other words, biomedical authors are much less likely to satisfy the triadic closure property than physics or mathematics authors.