Wiki connections

Protocol

Introduction

In order to explore the semantic area related to those selected Wikipedia pages about the Right to be forgotten, we used Seealsology tool, which allowed us to extract all the links in the See Also section producing a graph.

Step of the procotol

After selecting 5 Wikipedia pages linked to the topic of the research, extract the See Also urls from each of these 5 pages: on your wikipedia page, right click and select Explore Element, copy the div where the See Also are and then paste it on extractlinkurl.com. Then click on Extract all urls. In order to pinpoint the topic, read the See Also and highlight those that are actually inherent to it. Then, on Seealsology (labs.densitydesign.org/seealsology) put the urls of the 5 initial pages and of the See Also highlighted as relevant so that you can begin to explore the topic in a visual way. Put Distance=2 and start crawling (do not do anything else during the process so that it doesn’t freeze): by putting Distance=2 you are able to collect all the See Also (starting from your pages) up to 2 levels. If you put a higher distance, once you are able to import the graph in Gephi, you would get an almost incomprehensible graph, very dense and with up to 9000 nodes, making it really difficult to read. Once you have downloaded the .gefx file, open it in Gephi, selecting Graph Type: Directed, then you should put the following parameters:

  1. In Statistics, select Average Degree
  2. In Filter, select Topology, then Degree Range: remove nodes with degree=1 (2 -> max value)
    1. If step 2 doesn’t work and nodes with degree 1 won’t be removed, go to Data Laboratory, select all nodes with degree=1, right click, Delete All
  3. In Ranking, select the Nodes tab, then the “diamond” icon, then in the menu select Indegree, then Apply
  4. In Layout, select Force Atlas 2, with the following parameters: Scaling=0.35, Gravity=0.25, flag LinLog Mode. Start spatializing with Run (let it run until it stabilizes), then Stop. Flag Prevent Overlap, start spatializing again, run it until it stabilizes, then again Stop.
  5. In Statistics, select Modularity.
  6. In Partition, click the Refresh icon, then in the menu select Modularity Class in order to add colors to the nodes according to the clusters that have been created.
  7. In order to give the clusters that have been created the color you prefer, go to Data Laboratory, filter by Modularity, then pick all the nodes with the same class and then Edit Nodes by right clicking on them. Now you can choose precisely the color you want.

Metadata

Timestamp:
14/12/2014 - 15/12/2014

Data source:
Wikipedia

Tools:
Seealsology, Gephi