Collaborative notes taken by paricipants of the DH Summer School Switzerland:
Tutorial/Workshop Session 3: Martin Grandjean (Lausanne University)(@GrandjeanMartin): Introduction to Network Visualisation with GEPHI
Network Visualisation Tool: https://gephi.org/
Introduction to Network Visualisation
Import csv files (we don’t talk about the preparation of the data today).
File for Nodes (points)// File for Edges (connexions between points)
Edges have a source and a target
There are several centrality measures: By degree (number of connexions) / By closeness (closeness to the entire of the network) / by Betweenness (bridge part of the network) / Eigenvector (pageRank-style centrality – the more well-connected nodes are linked to you, the more you are central). And the more you are central, the more your neighbors are too. [a really good piece of work on centrality indices http://link.springer.com/chapter/10.1007%2F978-3-540-31955-9_3]
Examples shown during the workshop:
- Twitter during a #DH conference : http://www.martingrandjean.ch/colloque-dhiha5-les-digital-humanities-se-deploient-sur-twitter/
- GeoLayout plugin about european cities : http://www.martingrandjean.ch/archives-cartographique-geopolitique-megalo-urbanisation/
The complete tutorial to be able to analyse your facebook network : http://www.martingrandjean.ch/analyser-graphiquement-reseau-facebook/ Which is also a good tutorial to GEPHI if you avoid the part 1 about facebook, replaced by your own data or by the datasets prepared for the tutorial (data here: http://www.martingrandjean/gephi).