This week in class, we learned about how network analysis could help us to understand the relationship between discrete objects, and explored sophisticated DH projects that demonstrate how network analysis can be used as a form of visualization on various topics. One project that is particularly interesting to me is the Network of Thrones. This project provides the network for all eight seasons of Game of Thrones. The nodes are the main characters and the edges are weighted based on the interactions. The nodes are categorized into five main communities according to the story’s background. The networks display the strong triads in each community and the complex ties between communities. There are always a few central characters in each community, and all communities are connected in some way. The edges are weighted by centrality measures, which include degree centrality (how many people a character knows), weighted degree centrality (how many interactions a character has), eigenvector centrality (how many important people), page-rank (how many important interactions), and betweenness (does a character help connecting the network together – important connections). From these measurements, this project provides a reliable analysis of who are the true protagonists in each season, and how the relationships between the protagonists altered over the eight seasons. Even though I have never watched this TV series, by comparing the networks, I noticed the tremendous changes in the relationships between the protagonists over the eight seasons, which makes me curious about what kind of stories can lead to such changes. I’m impressed by how this project first turned vast narratives into data that can be analyzed and then converted back to graphs that summarize the whole story. It also combines network analysis with other forms of data visualization. For example, there is a character centrality diagram that measures the relationship between page-rank centrality, eigenvector centrality, and weighted degree centrality to give a sense of the scale of each character. I never thought that tv series can be analyzed in this way, which reveals the unlimited possibility of DH projects.
