This project is an uncompleted version of a sentiment analysis research project I completed as part of the 2018 UGA Digital Scholar Cohort.
The project takes three hard-coded inputs (different translations of a short story by Borges) and returns multiple analysis graphs of them. They can be used to track sentiment differences between translations, and I had hoped to observe drastic changes between them due to the time period each translation was created.
- Clone repo
- Edit each ".r" file to set graph size
png("Plots/Simple/Hurley/syuzhet.png", width = 1200,height = 700) - Once you run the scripts, it will create unsigned, signed, simple, and NRC emotion plots for each text using each available lexicon.
This project is unfinished. Future releases will allow users to set their direct inputs, choose which graphs to generate, choose a lexicon, and provide much a much better user experience.