Marco Toledo Bastos, University of São Paulo
Rodrigo Travitzki, University of São Paulo
Rafael Raimundo, University of São Paulo
This paper examines Twitter hashtags related to political upheavals in the Middle East, North Africa, Europe, United States and the Americas. We monitored the hashtags #FreeIran, #FreeVenezuela, #Jan25, #SpanishRevolution and #OccupyWallSt and compiled a dataset of over two million messages, including 791,968 retweets and billions of users interconnected as follower and following. We calculated the statistically significant correlations between retweets, mentions, followers, following and volume of messages. We also looked into the number of hashtag-related messages tweeted by single users regardless of their position in the network topology, providing a hypothesis to explain the emergence of Twitter political hashtags that is consistent with the practice of pamphleteering. The results show statistically significant correlations between retweets rate and number of tweets per user, but also low statistically significant correlations between retweet rate and the first-level network topology (follower and following network). These results suggest that instead of depending on user-hubs that act as opinion leaders, Twitter political upheavals are correlated with the intense activity of users who pamphleteer a cause or idea. We further examined this finding by comparing the hashtags network structure to the presumed communication pattern of pamphleteering, and we found a higher than average cluster of retweets sent and received by interconnected users. We also looked into the growing occurrence of Twitter accounts dedicated to activism and the high frequency of messages designed to increase the chances of a hashtag to spread through Twitter network. Lastly, we analyzed the text corpora of the hashtags looking for political charged concepts across the dataset. The results suggest that pamphleteering is a valid metaphor to the political activity of Twitter users, but also encourage further research using longitudinal networks to represent changes over time in the network.
Twitter; Protests; Hashtags; Arab Spring; Indignados; Pamphleteering