Implement a better profanity detection solution

[monkeytypegeorge]

Our current naive solution marks a lot of false positives. We need to consider existing technologies that might handle this better.