Between the COP 20 at Lima in December 2014, and October 2015, I was able to identify how a community of “influencers” became the most active and influents twitter users on the subject of climate migration. The 48 801 tweets about climate migrations (from 4.6 million tweets about COP 21) I collected thanks to the twitter analysis tool Visibrain during the COP 21 in December 2015 confirm this analysis.
The following social network analysis, created with the free software Gephi, is a data visualization of twitter users accounts. The bigger the node is, the more connections and centrality a node has with all the other nodes : it uses the PageRank algorithm, named after Larry page, founder of Google. The different colors represent the various communities. A community in social network analysis is a group of nodes that often mention each other.
We can see that the most influential community, in blue, includes most of the ten leaders I identified in my previous study. It shows that this community was able to progress from the 6th most important group identified by Gephi in the COP 20 tweets, to the most important community, without losing much of its leaders and members.
This influence can also be demonstrated by looking at the TOP hashtags and expressions from Visibrain : out of the hashtags used by the group of activists during COP 20 to promote a climate migration: #migration, #displacement, #disastermobility, #refugee, and #humanrights, only #migration was in the top ten most mentioned hashtags, with 106 occurences.
We can use the Tableau Software interactive dashboard to see how the mentions of these strategic keywords have progressed until COP 21:
The progression of the « refugee » and « migration » mentions in tweets is spectacular. “Displacement” and “Human rights” only get a few occurences but from December 2014 to December 2015 it is a 10 times increase.
If you want to make your own keywords insights I used the dataviz tool Tableau Software to create two interactive dashboards:
How does it work ? There is a time frame, a tweet map on the left, a tweet mention search on the upper right, and a most mentioned accounts view on the lower right. Most of these data are interactive : when selecting a specific time frame, the values on the other viz will change.
Selecting a specific location will also affect the tweet mention graph and the most mentioned accounts.
Finally, the tweet mention search allows you to compare 4 different keywords. The first keyword will also affect the most mentioned accounts graph. Selecting the graph corresponding to a specific key works will also change the localization value. If you want to see the evolution of all the tweets sent, make a blank search with the first keyword.
- The first one collects all tweets about climate migration gathered thanks to Visibrain from the 1st September 2014 to the 31st December 2015
- The second one includes only the tweets about climate migration sent from twitter accounts created in Pacific Islands