Who will die in the Game of Thrones?
As the new season of the Game of Thrones was approaching, I assumed I was not the only one interested in what it may bring, for instance, wondering which of my favorite characters are going to meet their ends, and which will live on to the next season. The aim of this mini-project was to shade some light on these matters and try to determine the probability of each major character’s death. To calculate that, I constructed the social network of the show’s fictional world using the episodes’ subtitles, calculated various network centrality measures of the major characters, and compared these values to whether they are dead or alive. For the comparison I used the well-known support vector machine technique making this project a combination of data mining, social network analysis and machine learning. As the show’s latest season’s finale is out, the prediction seems to be fair, while the method has caught the interest of many journalists publishing about this work on more than fifty news sites, such as Times Higher Ed, Futurism or GQ.
PhD candidate, Central European University
I am a Network Science PhD candidate at the Central European University with a quantitative background: Bsc degree in Physics, Msc degree in Biophysics, former projects on collective motion and social media analysis, and an internship as quantitative developer. My PhD thesis is about studying, quantifying, and modeling key components of career success on various artistic fields and comparing them to scientific ones. For this, I collect and analyze large scale data about motion picture, music, literature, science, etc. and develop a mechanistic framework to model the individuals career evolution on these creative fields. Besides the science of success, I am generally interested in data mining, data and network visualization, and social media.