Social Network Analysis

Modern marketing for films often involves utilizing social media, particularly Twitter, to reach a wide audience.  One way to engage a larger number of potential consumers is to create a hashtag and encourage its use in a particular way.  The upcoming Marvel film Captain America: Civil War is currently utilizing this exact marketing method on Twitter by encouraging the use of one of two hashtags: “#TeamCap” or “#TeamIronMan”.  Not only do these two hashtags encourage involvement by consumers, they also help to specifically market the movie, the plot of which involves pitting those two characters against one another.  This social network analysis looks at the relationships between and among some of those users who tweeted using one or both of these hashtags.

This social network analysis is comprised of a sample size of 2,000 tweets taken on March 21, 2016 from 1,642 Twitter users; it should be noted, therefore, that many users tweeted using one or the other of these hashtags multiple times.  This analysis focuses on tweets that have either or both of “#TeamCap” and “#TeamIronMan” as either hashtags or mentions.  Each point on the graph represents a single Twitter account. The graph is formatted in the Harel-Koren Fast Multiscale format, and is directed, meaning that all points are directed by arrows towards the source of the original tweet.  Lines between each user’s tweet represent follows, replies, or mentions between users. Self-loops, of which there are 224, represent tweets that are connected to themselves, likely retweets.  Blue lines represent tweets that featured only the hashtag #TeamCap; red lines represent tweets that featured only the hashtag #TeamIronMan; and gray lines represent tweets that either used both hashtags or otherwise did not indicate a preference towards one or the other.

Several interesting conclusions can be drawn from the social network graph.  First, it can be seen that the hashtag #TeamCap had a much larger number of lines than did the hashtag #TeamIronMan.  However, #TeamIronMan featured a much denser network, which can be seen from the few clusters of lines that exist on the graph; where #TeamCap was made up almost entirely of independent relationships, #TeamIronMan was almost entirely made up of connected users.  In general, however, the network had a very low density (roughly 0.0004).  This indicates that even though many users retweeted or responded to an earlier tweet, a large number of the Twitter users who utilized these hashtags were doing so independently, and not in direct relation to any other user.

Statistics-

• arianagrande: most followed (67628)
• youtube: most followers (60695078)
• citaveis: most tweets (354919)
• dj_gurj: most favorites (188420)
• Self loops are every tweet that is NOT a reply/mention
• Tweets in the network were tweeted over a 55 minute period from March 21st, 2016 21:56 UTC to March 21st 2016 22:51 UTC
• Vertices: 2046
• Unique Edges: 1964
• Edges with duplicates: 192
• Totals edges: 2156
• Self loops: 224
• Reciprocated vertex pair ratio: 0.092178649
• Reciprocated edge ratio: 0.004347826
• Connected components: 283
• Single-vertex connected components: 183
• Maximum vertices in a connected component: 1578
• Maximum edges in a connected component: 1735
• Maximum geodesic distance (diameter): 13
• Average geodesic distance: 3.395331
• Graph density: 0.000439763