What constitutes negative content?
The nature of communications varies greatly from one platform to another. That is to say that the way people interact on, for example, Twitter is not the same as the way they do on Facebook.
Therefore, we have different policies for different platforms.
Twitter is currently the only platform we provide our tools for.
The artificial intelligence is capable of recognizing sexist and racist tweets.
A tweet is sexist or racist if it:
  • uses sexist or racist slur.
  • attacks a minority.
  • seeks to silence a minority.
  • criticizes a minority (without a well founded argument).
  • promotes, but does not directly use, hate speech or violent crime.
  • criticizes a minority and uses a straw man argument.
  • blatantly misrepresents truth or seeks to distort views on a minority with unfounded claims.
  • shows support of problematic hash tags. E.g. “#BanIslam”, “#whoriental”, “#whitegenocide”
  • negatively stereotypes a minority.
  • defends xenophobia or sexism.
Adapted from the work of Waseem and Hovy (2016).