
119 - Social NLP, with Diyi Yang
NLP Highlights
00:00
The Modular Algorithm for Reducing Subjective Biases in Text
We usually only mostly process Wikipedia pages or the final outputs, but it's great to know about this process that's happening behind the scenes. So coming back to your actual task of devising or removing subjective biases in text, what exactly is the problem description here? You're given a sentence with subjective bias in it. And what do you do with it?Yeah, so the task here is actually a form of extenoration. So as the data set in itself, so we have the before and after. The input here is like bias this statement. And then the output here is neutralized attacks. Then the modeling task that we have is to generate this neutralized attacks, giving these
Transcript
Play full episode