Teradata is now using Big Data to predict the deaths in “Game of Thrones” – one of the most captivating works of literature and television experience. Clement Fredembach, a data scientist with Teradata’s Australia/New Zealand unit, used a combination of graph analytics and machine learning to predict the deaths of certain characters from one of the most unpredictable literary works. He discussed his predictions during a presentation based on his blog post at September’s Teradata Partners 2016 Conference in Atlanta.
The Process Behind the Discovery
Fredembach used information from the first five books in the series, uploading each chapter of every novel to his database which analyzed the text. He then created social networks using the raw data for each character which mentioned each of the 102 characters in the story.
Spoiler Alert! :
By the end of book 5 the “timeline” is back to normal, which makes it easy for us to prediction who dies in book 6 by looking at the state of the social network over the entire story.It is predicted that Jon Snow is alive and if his ground status is removed, then it is likely that all the other characters are to die.
The increased complexity of the story is well captured by the evolution of the social networks over time.