Tisane Labs, AI-powered text analytics solution provider, recently announced the Tisane APITrack this API. The API detects and classifies abusive text content across 27 languages. Abusive speech types include hate speech, personal attacks, unwanted sexual advances, criminal activity and the company will continue to add to this list.
"Trolls, bigots, harassers, and criminals made the Internet an unpleasant and at times dangerous place," Tisane Labs CEO, Vadim Berman, commented in a press release. "Now, with Tisane API, the communities online can automate much of the moderation process and even warn potential offenders before the post is published. Rather than producing a blanket statement and a floating-point figure, Tisane API pinpoints the actual instance of abuse and classifies the type of abuse."
The off the shelf API is cloud-based and RESTful. Those looking for custom integration can use an on-premise version of the API. The API includes a Language Model API and a Text Analysis API. The Language Model API allows for browsing language model data for common words, word-senses, and features. The Text Analysis API intelligently extracts meaning from certain words to detect abuse, sentiment, and more.