The “Network of Thrones” is a project created by Andrew Beveridge, professor at Macalester. In this project, professor Beveridge created several network graphs regarding the famous show “Game of Thrones”, visualizing the relationship between major characters in the show. Creating a separate graph for each season, he was also able to show the change of character relationship when the plot goes on.
The nodes are the character names, and the edges are the interactions between the characters, for example in the scripts it includes conversation, mentioning of another character, and two characters appearing in the same scene, appearing in a stage direction together, and another character mentioning Character A and Character B together. While in the books, interaction includes, two characters appearing together in the same location, two characters in conversation, one character talking about another character, one character listening to a third character talk about a second character, or a third character talking about two other characters. In conclusion, all interactions between characters were given their own weight and used to calculate the quantitative relationship finally draw the graph.
Specifically, professor Beverdidge used the following method. After compiling an exhaustive list of characters and their nicknames, he generated a network for the first volume, “A Game of Thrones,” by linking two characters each time they appeared within fifteen words of one another. Each link corresponds to an interaction (either direct or indirect) between these characters. And this result also showcases an interesting discovery, which is the relation change between characters over time. According to the professor himself, “This first volume clearly favors the Starks, but PageRank places Tyrion Lannister in second place, presaging the narrative expansion coming in the subsequent books. Meanwhile, the network penalizes Jon and especially Daenerys for their remote separation from the main action. “