The code in this repo was used in the data analysis in the following paper:
"Gender in the Disclosure of Loneliness on Twitter during COVID-19 Lockdowns" by Yelena Mejova and Anya Hommadova Lu
In submission to the Special issue of Frontiers in Digital Health in Virtual Presence: Loneliness, technology and the production of human (dis)connectedness:
It includes:
- Gender detection using Twitter user names via name dictionaries
- Removal of posts likely to be spam
- Temporal plotting of the posting volumes
- Content analysis using tokenization and Odds Ratio
- Creation of cooccurrence networks of select words
- Comparison of engagement metrics (likes, retweets, replies)
For more information, and about the original data (which cannot be shared due to the Twitter Terms of Service), please contact the first author: Yelena Mejova yelenamejova@acm.org