Expert Galleries: Moviegalaxies
For this installment of Expert Galleries, we're shining a spotlight on a set of network visualizations from Moviegalaxies, a website that enables you to explore the social connections among and between characters in films. Moviegalaxies is a project of Jermain Kaminski and Michael Schober, social network researchers. They've curated 8 visualizations in a gallery and taken us behind the project below. Thanks Jermain and Michael!
The Social Graph of Good Will Hunting (1998)
Moviegalaxies has mapped the social graph for 775 movies. How did you select the 8 that you included in the gallery?
We selected movies that demonstrate visually interesting graphs, represent different genres and are among the most requested on the platform. For the website, it was a trade of between the availability of the movie script and the visual representation of the graphs. Later we also took into account user requests and new releases. In future, we want to put more focus on user searches, user requests and graph complexity.
How did you order the 8 that you included in the gallery?
It's a mix of graphs that we find visually appealing, but also show the contrast of graphs between movies. In the end it's some of the graphs we looked at the most and found to be very fascinating.
How did you decide to start mapping the social graph for movies?
We both were working as students in the field of social network analysis at Cologne and MIT. The chances our paths would ever cross were quite low, however the day came where we exchanged some ideas and teamed up to create something that obviously didn't exist. We thought that movies would be a good vehicle to fascinate more people for social network analysis. Later, the idea developed towards a conference presentation, from the presentation to a website and then to now something more — a little like a snowball running downhill. Looking back on three years now, it's exciting to see that Moviegalaxies is used in some university classrooms around the world and has attracted interest beyond research.
What is data visualization allowing us to see in these movies that we might not otherwise see?
We think that every movie has a beauty that is hidden behind the effects, sounds, cast and story. This beauty is the social network design. Also, movies might have more in common with our own life than we think. For example — think about the social network designs of successful movies such as Star Wars, Lord of The Rings or James Bond. In many cases you will likely find a network with at least 3-4 subclusters of networks that the main characters are interacting with. Now take a look at your own life: You will likely have 3-4 clusters of networks, such as family, friends, work and university. If you would draw the connections between the nodes in your personal life and compare that to one of your favorite blockbuster movies, you might be surprised about the similarities — at least structurally. From this perspective, one could think of something like "network empathy" that viewers might develop.
What is the diameter of a network and what does it reflect for a movie?
The diameter is a term originating from social network analysis and describes the length of the longest path between connected nodes (characters). The idea of a network diameter is to index the extensiveness of the network, how far apart the two furthest actors are. In many cases, a high network diameter reflects a complex and wide ranging network with many sub-clusters (Babel), while a low network diameter is present in movies with high centrality and a central figure (Forrest Gump).
The movies in the gallery span from 1977 to 2008. Have you observed the networks changing over time?
While we are improving on the scientific substance, we observe that the networks in movies became more complex. At least, this observation resonates with how our modern society is organized: in complex networks.
What is the most important insight you'd like viewers to get from this gallery?
We like to get our users interested and fascinated to discover movies in a different way than they are used to. We'd like to catch the eye of the visitor with a beautiful and clear representation of their favorite movie. We'd like them to recognize the structure of the movie and the main characters straight away. We find it fascinating to talk about the story of a movie while looking at the graph, because you immediately recognize the structure between characters. If you had 30 seconds to explain the story of a movie to another person, Moviegalaxies would be the best place to support you with this. We will continue improving the idea and add some interesting features in the near future.