The concept of community detection as measured by the interaction patterns of social media users who form groups around a particular event is the focus of this paper. In this particular case, the authors use two real-world events as their case studies: a series of Eurogroup meetings and a local TEDx event. Twitter conversations about these events serve as the database from which an evolving community forms as a result of the members of the group being more densely connected to each other, sharing common properties when compared to others in the network.

Microblogging sites were used to track these events as well as Twitter feeds. The authors use a graph to track the dynamic evolution of the formation of communities and to show the intensity of this relationship. They connected blog posts based on their textual similarity as well as the similarity of their temporal activity profiles. The impact of the general topic on the various intended audiences could be deduced from these sources. Users from various disciplines rely upon them to share news/opinions. The same was done for mentions, replies, and retweets generated in Twitter.

Aside from graphs, tables and equations were used throughout the paper to show how the data sets were generated and how they summarize information regarding “time duration, total number of tweets for the period, [the] number of interactions among users in the collected tweets, and the number of users.” Active communities are identified and classified as small when there are less than 50 members, medium with between 51 and 200 members, and large with more than 200 members.

The value of such information stems from its usefulness to applications in marketing when one is trying to determine a campaign’s success or to monitor a particular brand; in “politics and public affairs [when] estimating the approval/disapproval of a new policy or tracking uprisings of groups of people[; and in the] news media and event organizing.” Thus, the work contributes to the formation of a “generic framework for community tracking” and the identification of communities at the local and global levels; and serves as an approach for global event, intra-event, and cross-event analysis.

Except for the difficulty with the phrasing of the material in this paper stemming from the original language of the authors, the paper shows how case study analysis can help in the understanding the formation of communities within an electronic network.