The project I explored is Movie Galaxies. It’s a website that visualizes character interaction networks in films, offering a unique perspective on how characters connect within a film (answering the goal of the project). It engages with the audience of films and those who study film and media. It clearly shows the relationships between characters and helps people better understand the plot (answering the academic fields the project engages with). The website has collected information on 772 movies from various periods. The latest one is The Surfer King, released in 2020, while the oldest film in the collection dates back to 1915, The Four Feathers. All the films on the website can be sorted by genre, and you can organize them by popularity, title, or release date. For each film, there’s a network showing the interactions of all characters. It uses a chart to describe the frequency of interactions between characters, where greater frequency results in a larger block representing the character. Moreover, it compares the data of one film with others in the database and provides statistics like the number of characters and the number of edges in the network.
Sources: The network graphs and metadata can be found on Harvard Dataverse. Most of the information used in Movie Galaxies is about characters and their conversations with others in films. As described on the website, the publishers developed a movie script parser to measure character connectedness based on their co-appearance in the same scenes, with each co-appearance counted as one degree unit per scene.
Processes: Data is collected using the method mentioned above. The connectedness of characters represents the edges, and the characters themselves are nodes. Then, JavaScript and programming languages like Python are used to visualize the data in the form of network graphs.
Presentations: Network graphs are used to display the connections between characters, while treemaps are used to show the frequency of character interactions.
My question is that the data used have minor errors, is it necessary and important to be fixed for a digital humanity porject?
