It's still imperfect, but sentiment analysis and flagging known slurs might be a good start. You can roll-your-own, or use a third-party API (whatever the impl

It's still imperfect, but sentiment analysis and flagging known slurs might be a good start. You can roll-your-own, or use a third-party API (whatever the implications might be).