I really enjoyed this midterm project, it required the use of many skills that I have developed throughout this course. This is a summary of the process of taking data from X-men Uncanny Covers and processing it to present two forms of data visualizations on the characters in the covers.
Sources
I decided to use the X-men Covers dataset. This provided a wide variety of information about the covers of several successive issues of X-men comics. There was data on who the cover artists are, the characters seen on the cover, any narrative captions there were, and which character is speaking. Though there were a lot of data columns, there were a lot of data missing from each row. For example, not many covers had dialogue.
Processes
As a result of this, the main type of data that was usable was the connection between characters in different covers. Therefore I wanted to show two types of visualizations, the connection between characters, i.e. how connected is any character with the others, and also how frequently the characters appear. I decided to use Voyant tools since the presentation was data visualization. Firstly, I did a dry run by using the data without any cleaning to see how it would affect the graphs. I noticed two main problems. Since the names of the authors appeared often, it was being treated as a character, also some words that appeared in dialogue would have a high count and appear in the visualization. To clean the data I used excel, I deleted all columns that had dialogue or text since I wouldn’t need it. Secondly, I deleted the names of authors after it was mentioned once.
Presentation
After cleaning the data, I was able to obtain a Plot and a TermsBerry shown below.
Plot of the relative frequencies of the 5 most popular characters over 10 segments
Each colour represents a diffrent character.
Terms berry Showing the connections between all characters
Hovering over a character(which becomes highlighted in green) and will allow you to see every other character they were in a cover with(highlighted in red) and a count of how many covers they have been in.
Significance
One main difference between DGAH and data science that I learned from this process is that there is often extra information in the datasource in DGAH than datascience. Since the data obtained in DGAH projects is processed digitally, not all of it is needed for a given form of presentation. As seen in this project the author’s name would be a vital form of information but when working with Voyant-Tools, it was not needed and could be removed.
