Automatic detection of problematic content in text, including attempts to lure users off the platform and establish content, profanities, mental issues and suicidal ideation, and more.
Built from scratch for social media content, Tisane handles misspellings, algospeak, bad grammar, jargon. It operates with social content is a manner that is more granular than any other NLP platform.
Tisane problematic content detection supports 35+ languages, including slang and obfuscated text (adversarial text manipulation).