https://project-archive.inf.ed.ac.uk/ug4/20191374/ug4_proj.pdf
In this project, we want to know if emojis are used differently by different personas.
The main categories of analysis are:
- Gender: Males, Females
- Race: White, Black, Asian, Hispanic
For each group of persona, the following analysis is applied:
- Percentage of usage of emojis
- Top used emojis
- Distinctive emojis
Machine learning models are built for gender and race detection of users using emoji.
Work steps contain the following:
- Collecting millions of tweets with emojis
- Trying different methods for gender and race estimation of Twitter users
- Performing statistical analysis for data
- Building user gender and race detection models using only emoji usage in users' timeline