Technically it is interesting to identify topics and classify speech but I doubt any underground forum would be interested in applying this. I wouldn't want thi

Technically it is interesting to identify topics and classify speech but I doubt any underground forum would be interested in applying this. I wouldn't want this for my forums either. > Given the importance of the problem, tackling hate speech has become the main target of many studies I think it is neither important nor constructive. On the contrary I think this will only deepen conflicts as people perceive such rules to be biased, which they objectively almost always are, not only in this work. I think almost all conclusions that are drawn here are wrong and lack evidence.