I chose to look at Movie Galaxies which is a digital humanities project that explores the interactions between characters throughout a total of 772 movies and series. For each movie and series, the character interactions are shown in a as a interactive network which allows the user to hover above any character to get more information and see which other characters that character is linked to. The project consists of many interactive network visualizations, each one being of a different movie or series, with the title of the movie and the year it came out. Each visualization looks something like the one below, which allows for easily visualizing what the character interactions in a given movie looks like.
The goal of this project is to provide further information on the interactions between characters and how these interactions are interlinked throughout the movie. I can see this project being most helpful for a cinema studies class due to its entirety being about movies and series. I feel like this project could be expanded to be used for books or plays and be used in the field of literature.
The data used for this project is each movie/series’ screenplays. The source of the data is the Harvard Dataverse page, which created a a breakdown of each screenplay and each scene to determine which characters appear in the same scene to find the connections between characters.
I couldn’t find specifically what processes were used, but the Harvard Dataverse page explained that each interaction is measured as one degree unit per scene (so the more interactions characters have, the more degree units will be listed for each character). It also stated that the data includes individual network graph data in the format of Graph Exchange XML along with descriptive statistics on measures. Below is an image from the Harvard Dataverse page that describes the data in more detail.
The presentation for this project consists of compiling the data for each movie into individual graphs, with each character being a node with lines to connect it to other nodes (characters) to display the character interactions. Each graph is interactive and allows the user to click on nodes for more information about that character.
One question that arose as I was breaking down this project was that since everything is already setup, if it would be easy to add a section to the page for books/novels? if they could use the same processes to upload the books and break down the text to find character interactions in each “scene” of the book in order to create similar visualizations for hundreds of books? I think this shouldn’t be too hard as they already have the project and processes set up. The only thing that might need tweaking is the process of breaking down the scenes as screenplays and books differ in their format.
