Context
Recently on Twitter, users have taken to sharing screenshots of tweets instead of retweeting the original tweet. There are a few reasons to do this: 1) it’s stealthy and avoids unwanted confrontation (the author of the tweet is not notified of the share as with a retweet) and 2) it circumvents the Twitter popularity metrics, preventing the amplification of the tweet in other’s timelines. As such, this is often used when a user wants to disagree with a tweet or attack its author indirectly.
This project studies this in the context of attacks on human rights organizations, specifically the ICRC, which are often threatened by manipulation campaigns and online abuse.
Steps
- Develop a data collection methodology to capture screenshots of Tweets
- Curate a dataset of tweets with tweet screenshots
- Find the original tweet and compare its retweets to the screenshot tweets
- Classify the tweets to understand why the tweets were shared as screenshots and not retweets
Requirements
- Proficiency in Python
- Some experience with Data Analysis
- Some experience with Machine Learning
- Some experience with Natural Language Processing