Sift Ninja Features
The quickest, cleverest and most accurate filter for unmoderated content. Language isn’t black and white so your filter shouldn’t be either.
The Smartest Content Filter on the Internet
Sift Ninja at its core is a text classification system that identifies and labels online content. It understands patterns of behaviour like sexting and racism. Powered by Two Hat’s Community Sift, you benefit from our robust and powerful unNatural Language Processing capabilities.
Included in Sift Ninja are our top profanity topics like vulgarity, sexual harassment and racism are detected and assigned a risk level. Messages are flagged with this metadata (topics and risk levels) and sent back. Your platform then decides, based on its own rules, what to do next.
Our simple, self-serve and easy to use API auto-filters, and removes profanity from your community while allowing a range of expression that fits your community.
Use Sift Ninja just about anywhere.
With our API, you can use Sift Ninja anywhere you need to protect social interactions. Games, apps, mobile, bots, etc.
Text classification — More advanced than blacklisting & whitelisting.
- Real-time dashboard reporting
- Install on multiple platforms at once
- API-based service works with any platform
- Instant filtering based on a sliding scale — from low risk to severe
- Best-in-class technology for text classification
- Fast & simple setup
What can Sift Ninja Protect Against?
Vulgarity is defined as any comments that are lewd, crude and use profanity in a general sense but not inherently targeted at a person or a group of people.
Sexting refers to any attempt by users to send and/or receive sexually explicit messages or discuss child pornography.
Hate Speech refers to any speech that attacks a person or group on the basis of of personal attributes or identities.
Personal Identifying Information (known as PII), is information shared online that can be used to identify someone in real life.
Handles l33t speak and filter bypass tricks.
Our system automatically looks for hidden letters and numbers and even characters from other languages. We have also built out thousands of unique rules to account for subversive manipulations.Start Filtering Now