The ideal technique to figure out how language is being made use of negatively to lead to harm is to crowdsource it from persons who have been harassed, and then experiment pulling tweets from Twitter using facts-accumulating bots and analyse the benefits on your own. |