We are just two weeks from the open Coursera MOOC "The History and Future of (Mostly) Higher Education." The course is an important piece of the FutureEd Initiative that includes dozens of classes, research projects, and informal groups spread around the world.
You can check this list to get detailed information on related courses and projects. But be prepared. This list is long. It is so long that we thought we should visualize the classes and institutions affiliated with the FutureEd Initiative by modelling the keywords and plotting the results in a word cloud.
I pulled the data from the roster into R and used topicmodels to estimate a CTM models using the variational expectation-maximization (VEM) algorithm. Below you find two similar but different results from the topic modelling. These images are a good example of how different iterations can produce diverse term commonalities.