A browser extension to help users avoid potential psychological triggers on the internet. This application is a chrome extension that parses the webpage content and detects potential triggers and blurs the text with a <TW: reason> and a button to unhide the blur if the user wishes to read it anyway.
These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.
Google Chrome*
*version to be updated
- Go to chrome://extensions/ and enable developer mode
- Click "Load unpacked extension" and select root folder to add extension to chrome
-
Milestone-1: Detect and blur sexual assault related triggers from Google news Headlines
- Set up Chrome Extension
- Fetch headlines on page load completion
- Detect potential triggers in the fetched headlines
- Update DOM element to blur text
- Allow users to add their own triggers
-
Milestone-2: Context based detection of Triggers- Using Machine Learning
- Set up Flask server
- Write REST API to accept POST requests with dummy payload and respond with a JSON
- Fetch data and label them for building the ML model- Triggers and Non-triggers (Third person news reporting trigger data is not available)
- Build a Logistic Regression model to classify Triggers
- API tests
- Integrate with Chrome extension
- Optimise request-response time
-
Milestone-3: Extend supported content beyond Google news headlines to full length articles
-
Create issues on github for pending action items
-
Upload Specification document
- Divya Sivaram