WatchTower
WatchTower is a wrapper around the Twitter API designed to help users block undesirable content such as hate speech, violent extremism, misinformation, and more.
The inspiration for watchtower came two fold; first from my PhD research into identifying and predicting ideologically motivated cyber crime and secondly from my interest in building tools to support users. WatchTower builds off these areas by providing a platform for Twitter users to perform mass-blocking of violent extremist users.
Bubble Check-In
Want to know what your social media bubble actually is? Or maybe you want to check-in-on a friend who you worry is experiencing hate online? Bubble Check-In helps you get a picture of the type of content a user is experiencing on Twitter! 🐦 Bubble Check-In takes a Twitter username and provides several outputs, graphs, and figures detailing their exposure to hate speech and negative sentiment online.
Releases
PINPOINT
Pinpoint is a set of Python tooling that wraps around the creation and use of a binary classification model for identifying far-right extremist content in social media posts (specifically on the Parler social network).
BLINKERS
Blinkers is a simple UI based text editor that allows for filtering sensitive, violent, and radical content limiting the exposure to researchers. Blinkers currently supports the following features:
- Open, Save, and Modify text files just as you’d expect from any text editor
- Highlighting multiple words in a file
- Change the saturation of non-highlighted text, limiting exposure
Obfuscate_Protect
A tool for quickly obfuscating sensitive, violent, or extremist text to protect the researcher or analyst viewing the content – while keeping it in a unique form so that patterns can still be observed.