1363. sredin seminar: Vladimir Batagelj: Drilling into Erasmus learning mobility flows between countries 2014-2024

Drilling into Erasmus learning mobility flows between countries 2014-2024 Vladimir Batagelj

At the bottom of the Erasmus+ page Learning mobility projects data visualization, there is a chart “Learning mobility flows since 2014”. It allows for interactive exploration of mobility flows between countries. It looks nice, but the insight gained is very limited. It also offers the possibility to download network data.

In our presentation, we propose alternative exploratory views on this network. The Erasmus network has 35 nodes (countries). It is very dense - almost all possible links are present, and the range of link weights (number of visits) is very large (from 1 to 217003). The range can be reduced by using a monotonically increasing transformation that preserves the order of the weights. Traditional graph-based visualization is unreadable.

To evaluate the position and activity of countries, we first determined Ps cores and Hubs and authorities. To gain insight into the structure of a dense network, it can be reduced to a skeleton by removing less important links and/or nodes. In our case, we have determined a 1-neighbors (only the strongest link in each node is kept) and 2-neighbors (+ the second strongest link) subnetwork, and the Pathfinder skeleton.

Another option is to use a matrix representation of the entire network. Using the ”right” order of the nodes can reveal the network structure as block patterns in the displayed matrix. Such an order is usually determined by clustering. We also applied Balassa normalization (log(measured/expected) visits) to the Erasmus network.

The obtained results will be presented at the seminar.

Seznam preteklih seminarjev je na seminar-sreda.

PS. Kdor bi rad kaj povedal na naslednjih seminarjih, naj mi sporoči naslov teme in doda kratek povzetek.