Outils logiciels pour les cours Paris II

Cours Paris II

Stages/ Thèses

edit SideBar


Analysis of Social Graphs with Gephi

Prerequisite :

You need to have Gephi. There are two versions 0.8 and 0.9 and they differ in their way to interact with Gephi. Below is the configuration of Gephi 0.9.

  • Step 1: Outils, Modules d'extension: Graph Streaming et Twitterstreamingimporter
  • Step 2: Twitter streaming, credentials. Suivre le lien, identifiez vous sur Twitter, créez une application, copiez le Consumer Key (avec le secret) et le Access Token (avec le secret).
  • Step 3: Twitter streaming window, add Keyword
  • Step 4: Streaming window: Master Server doit passer de rouge à vert (bouton droit)
  • Step 5: Twitter Streaming window, Connect (en bas)

Une séquence d'arêtes doit apparaitre dans la fenêtre principale.

If not, go to http://gephi.github.io/ and download Gephi.

Important : Gephi works on Windows, Mac and Linux but only with Java 6 or 7 (not 8)

If you have an installation problem with Gephi go here http://gephi.github.io/users/install/#gephi-does-not-start.

Tuto & Tools for collecting data for social networks

Data Format accepted on Gephi :

Figure 1

User Interface :

Figure 2

Plugins :

Before we can use Gephi with streams, we need to install additional plugins.

Go to toolbar, select Tools / Modules / Modules Available

Select :

  • Graph Streaming
  • Chinese Whispers Clustering

and click to install. After installing, restart Gephi

Or manually method :

Go to the menu Tools > Plugins > Downloaded. Click on Add Plugins…, and select the NBM or ZIP file you have just downloaded. Click on Install, and follow the wizard.

https://marketplace.gephi.org/plugin/graph-streaming/ https://marketplace.gephi.org/plugin/chinese-whispers-clustering/

Step 1 :

For importing a dataset, click to [File / Open] on Gephi menu or double click on the file (with good format for Gephi)

If you use Naoyun, go to the Streaming tab and start the Streaming server.

Step 2 :

You should see a really compact cloud of nodes (more or less dense depending on the size of your network and its connections).

Select “Force Atlas 2” in the drop down menu of spatialization and click to Start.

Figure 3

This property is very useful to interpret a graph:

  • Groups of nodes can be spotted immediately,
  • Relations between groups are also visible.

Step 3 :

The degree of a node is the number of relation (edge) it has, to display node size per degree : Ranking -> Nodes -> Choose a Rank Parameter Degree and click « Apply »

Figure 4

This window offers you four options for both the Nodes and the Edges. Enabling you play with the color, the size and the labels display.

Step 4 : Composantes Connexes:

  • Statistiques: connected components
  • Fenêtre Aspect: Noeuds, Partition
  • Faire varier la taille des noeuds, par degré: Fenêtre Aspect: Taille, noeuds, degré

Step 5 :