bias-monitor VS their.news

This is my project, check it out in my Reddit history… not super pumped about the repost claiming it’s yours. Please at least give a link to the project: GitHub

Good thinking, this model was trained on 3 different datasets (found here) and they included data from both inside and outside of America.

Can’t trust anybody… but the project is all open source. So you can go take a look at the code for yourself on GitHub here. I also have a writeup on the README for how the data was labelled for the ML model.

Exactly. The biases are per news source and actually constant: https://github.com/Hadjimina/perspectiveNews/blob/0c4c6a932a...

I'm sure they are eventually updated, but making a general judgment per news source to show individual articles means it will occasionally be wrong and misleading for the reader. The name alone, "their media", suggests an us vs. them mentality that we could use less of these days.

For once I think machine learning would be suitable for this. Have something like GPT-3 process individual articles and produce a bias number. I think this is how https://www.improvethenews.org/ works, and at a quick glance, does a much better job at classifying.

But honestly? We shouldn't bother. Consume every new information with a critical mind, and don't expect editorials to do the thinking for you. There are no bias sliders for social media posts, where most information comes from these days, and we need to teach critical thinking instead of picking what we want to agree with or not.