Interdisciplinary Teaching at UNCA - a Network Graph

UNCA prides itself as an institution that values interdisciplinary teaching. The network graph below is an attempt to visualize this. The nodes () represent academic departments. There is a link between two nodes () precisely when there is at least one instructor who teaches in both departments. The size of a node is indicative of the number of instructors in a department and the width of a link is indicative of the number of instructors in both departments. You can hover over a node or edge to get a bit more information. You can also drag the nodes around a bit.

Comments

• The data for this visualization was all scraped from UNCA's schedule pages for this academic year - Fall 2017 and Spring 2018. That process yields prefixes, like "MATH", "STAT", or "ARTH", rather than departments like Math and Art. The lumping of just over 50 prefixes into 30 departments and then those departments into 5 divisions was done by hand. I think it's pretty good but, possibly, open to debate.
• Engineering was not included for the simple reason that the web pages where the data was scraped from list exactly one instructor for every single engineering course. Graduate classes were also excluded.
• We want more than a simple visualization; we'd like to quantify the degree to which UNCA values interdisciplinary teaching. With this data in hand, there are a few ways to do this. At the simplest level, we could simply observe that UNCA is employing 347 instructors this year and 104 of them are listed as teaching in more than one department. Thus, nearly 30% of instructors are "interdisciplinary".
• On a somewhat deeper level, we could use any of several measures of connectivity that arise in the theory of network graphs. One such measure is called average node connectivity. I computed this for the network to be about 3.2. W00t! Of course, it's hard to interpret this without some context.
• The layout is achieved by a force directed algorithm. All nodes repel one another like electric charges. The links between the nodes create an attractive force and the layout evolves until an equilibrium is achieved. Of interest for us, is that the apparent centrality of the interdisciplinary and core courses arises from that layout process. I did not place them artificially.
  Indeed, there are a number of concrete measures of node centrality in the theory of network graphs. Having computed several of these, I consistently found the Interdisciplinary department to be the most central, followed immediately by the three core departments.
• As fun as all this is, a glaring weakness is the complete lack of context. I'd love to be able to visualize similar data for related institutions. I'm just not sure how much of a challenge it might be to get that data though.