Researchers at Central European University in Hungary have developed a machine-learning algorithm that can predict which characters on "Game of Thrones" will die next.
The researchers quantified each character's social significance, and then entered those figures into the algorithm.
The team built a network of the fictional world's social system based on how often characters interacted with each other by pulling data from the show's subtitles, providing almost 600 scenes' worth of information about how often characters appeared in the same scene. The data illustrated which characters had the strongest ties to other characters.
After calculating the most important characters, the researchers related "network position to survival."
Of the 94 characters considered in the study, 61 had already died, providing information about their social position and level of importance.
The researchers found the model accurately predicted the fates of about 75% of the characters that were analyzed.
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA