VADER: A “Sentiment”al Journey Through Drag
Elena C -
Hey everyone,
So after conducting the last of my interviews I went ahead and made a slight change to some of the coding process. I wanted to use a different system: the VADER (Valence Aware Dictionary for Sentiment Reasoning) code. I used this because it is more helpful with social and political commentary. It is also important to note that I was able to account for punctuation with VADER technology as well, which is really helpful because the interviews were all recorded via Zoom, and therefore I was able to use facial expressions (based on my own understanding) mixed with their phrases to add excitement or pauses into the transcripts if they were not caught. This allowed for a better understanding for contextual texts to be detected such as sarcasm. Below is an example of one of the graphs I made from data from one interview. The name is whited out for now because I am still receiving confirmation on which drag performers would like to stay anonymous. While there were more questions asked such as greetings and getting to know each drag performer, I cut those out of the sentiment analysis and only analyzed those pertaining to political and social questions about drag. The questions and their respective numbers will be visible in both the paper and presentation but for now I just wanted to give an example as to how the data will be digested. This also means slight adjustments will be made to my ML models due to a small change in data and those will be available to see next week.
