Project Summary
George RR Martin’s ‘A Song of Ice and Fire’ is one of the highest grossing fantasy/fiction series to ever be published. Spanning multiple realms and littered with an endless number of colorful characters, its visceral tales of greed, ambition, love and betrayal have created millions of fans world-wide (me included). HBO took the series to the next level adapting it into the equally successful ‘Game of Thrones’ TV drama in 2011.
As the eighth and final season of GoT is just around the corner in 2019, I decided to do an exploration of all the character interactions in the original ‘A Song of Ice and Fire’ saga using network science techniques. Using these methods we can work out who is the most connected and establish which characters are particularly important to the storyline.
Using node and edge data from 'A Song of Ice and Fire'
Node and edge data was created using the entire ASOIAF anthology where nodes represent characters and edges represent a narrative connection between characters.
Graph modelling using Networkx
Networkx was used to model the network. Degree and Betweeness Centrality were used to quantify and make sense of how the relationships in ‘A Song of Ice and Fire’ are structured.
Interactive visualisations with Plotly
Plotly was used to create interactive visualisations of the network alongside stock plotting in Seaborn.