This is a preliminary result of the network of retweets with the hashtag #jan25 at February 11 2011, at the time of the announcement of Mubarak’s resignation. If you retweeted someone, or has been retweeted, it is possible that your username is one of these tinny points (or maybe a bigger one?).
To collect the network data, I used the Gephi Graph Streaming plugin, connected it to a Python web server I made myself. This web server works like a bridge, it connects to the Twitter Streaming API using the statuses/filter service and converts the users and retweets to nodes and edges in a network format that can be read by the Gephi Graph Streaming plugin. Nodes are twitter users, and links appear between the nodes A and B when B retweeted a message of A containing the hashtag #jan25.
The static network visualization is just the final result of about one hour of data collection. It is a dynamic network, and it’s possible to get much more information from the collected data. For example, before the announcement, there were few nodes and edges, sparse in time. But when the announcement arrives, a boom of retweets appears on the network. A video with the flow of retweets is available on YouTube. It shows the dynamic network construction during the hour of data collection, compacted in less than four minutes. During the collection, I run Gephi with the Force Atlas layout just adjusting some parameters from default: repulsion strength to 2000, attraction strength to 0.3 and speed to 10.
I was very lucky to get this data. On February 11 afternoon I was testing the Python server that works as bridge and connected to Twitter. I tried some interesting hashtags to see it working, and at the moment #jan25 seemed to be an active hashtag. I let the application run for some time, adjusted some parameters for visualization, and at some point there was a burst in the activity. I didn’t understood what was happening until I checked again my Twitter account and realized that the Egypt’s vice-president had just made the resignation announcement. After it, I proceeded collecting data, and the final result was this network. It was very interesting to see, in real time, the exact moment when Tahrir Square, from a mass protest demonstration, has been transformed in a giant party, and the burst in the Twitter’s activity. It was like covering in real time a virtual event, a big event that was happening in the Twitter virtual world.
After playing with the data, I found that the data I got through the Twitter Streaming API is only approximately 10% of the total. I’m now working to recover all data and hopeful soon I can make available the full graph of retweets.
André Panisson / www.
This work is part of a research project involving the Computer Science Department of the University of Turin (www.di.unito.it), the Complex Networks and Systems Group of the ISI Foundation (www.isi.it), and the Informatics department of Indiana University (http://cnets.indiana.edu/).