Profanity > profanity-check

Uses a linear SVM model trained on 200k human-labeled samples of clean and profane text strings.