Friday, September 11, 2015

Relationship of characters in Game of Thrones




Game of Thrones is an American fantasy drama television series, which has attracted exceptionally broad audience from home and abroad. The series is an adaptation of A Song of Ice and Fire, George R.R. Martin's series of fantasy novels. 

As the above banner points out, in Game of Thrones, you win or you die.  In the first four season of the series, the death toll of characters is 456. Characters appear, get killed and are forgotten by the audience. Due to the large amount of characters appeared, it's quite normal for audience to get confused by the relationship among characters.

Thus, the TV series has become an interesting topic for data visualization. There are majorly two different ways of illustrating how these characters are connected, through the traditional hierarchy diagram or social network analysis diagram.


Hierarchy Diagram 


The above diagram shows a clear relationship within each houses through marriage or birth. The reader can easily see who and who are related by blood. The author also added other relationship such as servants, killer and rivals to showcase the relationship between major characters. 

The family tree like diagram can easily show the hierarchy of characters, and readers can get a basic idea of the relationship between major characters. However, there are over 500 characters in Game of Thrones. It would be hard to illustrate the relationship of all characters in one single diagram.


Social Network Analysis 



This is an example of the social network analysis for Season 1 of Game of Thrones characters. In total, there are 120 characters identified. Each character is a node, and they are connected if the characters ever have some interaction, which includes verbal, physical or gestural actions. The degree of each node represents the "air time" of the character. Higher degree nodes will have a bigger circle in the diagram. Without actually watching first season of Game of Thrones, viewers can easily see that Eddard Stark is the most relevant character in first season.




Here is another example of network analysis of characters relationship based on the novel. In addition to the information shown on the previous diagram, it adds "kill" and time to the network. The Link between nodes means a "kill" instead of an interaction.


Learning and Thoughts

The following table compares and contrast the pros and cons of the two methods used to visualize the show,Game of Thrones. 

Method
PROS
CONS
Hierarchy Diagram
Ø  Shows hierarchy of characters
Ø  Very clear of specific relationships
Ø  Limited size and quantity
Ø  Only surface facts are revealed
Network Analysis
Ø  Reveals deeper connection between characters
Ø  Easily sort out centralized nodes (not necessarily the ones on top of the hierarchy)
Ø  Size and quantity of nodes are not limited
Ø  Not clear about specific relationships between nodes
Ø  No hierarchical view (specially important to understand the show)

Each method excels in their own domain, and has its disadvantages. However, it is quite pleasant to see that when used together to analyze this TV show, they complement each other. The Hierarchy Diagram provides great details about each relationships and shows a clear ranks of the social status of each characters, while the Network Analysis revealed deeper connections between characters. I would say that the Hierarchy Diagram helps audience understand the characters better, but the Network Analysis would probably prepare audiences for future surprises in the show. 

Our team are amazed that text mining and social network analysis can provide a new perspective and understand of the classical novels as well. For example, Dream of the Red Chamber, one of China's four greatest classical novels, can be interesting to analyze. The book has more than 40 major characters and over 500 additional ones. Apart from complicated relationship between characters, only two thirds of the book were confirmed written by the original author. Social Network Analysis probably can bring new perspective into the research exclusive to this novel. The examples above really changed our team's views about network analysis. Not only could network analysis bring benefit to scientific researches, but also it could complement the entertainment of our life. 

After all, it is quite interesting to see what we have learned in class can be used to help us understand the complicated relationship. At the beginning of the class, when first heard the theory "network", we automatically linked it to graphs with nodes. How excited could that be?! The answer turned out to be, "Super excited"! As we slowly dip our toes into network analysis with the guidance of Dr. Ram, we started to appreciate its beauty that it can fits into every aspect of this world. It has been applied in many scientific and social disciplines including machine learning, particle physics, the internet, biology, supply chain, social networks, event predictions and etc. With rapid technology advancement and data overflow, network analysis is becoming easier and more accessible to more and more people. This also enables each individual to make better and faster decisions on the run. 

Stay tuned for more insights about Big Data Analytics!




Team Big Bear,

- "Think Big, Like a Big Bear!"








Reference: 

https://en.wikipedia.org/wiki/Game_of_Thrones
http://hauteslides.com/2011/05/game-of-thrones-infographic-illustrated-guide-to-houses-and-character-relationships/
https://en.wikipedia.org/wiki/Dream_of_the_Red_Chamber
http://www.washingtonpost.com/graphics/entertainment/game-of-thrones/
https://medium.com/@nscmnto/game-of-thrones-beyond-the-power-struggles-ea4567451e67
https://en.wikipedia.org/wiki/Network_theory