AI / Machine Learning Models
Leverage AI & Machine Learning Models to scale your moderation decisions
Text, Image & Video Models
Dozens of powerful models to review your content on multiple vectors, to give you more info to make and automate your moderation decisions.
There are models that detect both destructive/negative conversations, as well as those that are constructive and positive.
Use this information to understand the general climate of the conversations in your community or to automate decisions on different types of content.
What are Machine Learning Models?
Machine Learning Models are programs that can find patterns or make decisions from a previously unseen dataset.
Machine Learning Models used in moderation are trained on very large data sets, usually vetted by experienced moderators who have rated the program’s decisions repeatedly, and used that info to retrain the model until the model is able to give a reliable rating on its own.
Tunable to your brand
Incorporating any moderation toolset takes time to integrate and calibrate to your company’s brand guidelines, The ratings from the various models you choose to employ can be used as additional information for your moderation team in their manual reviews, or you can automate based on certain thresholds you determine.
- Are you a kids/family focused brand with a zero-tolerance policy on negative content? - Turn up all your dials
- Want to keep out the truly bad stuff, but ok with a bit of colorful language? - Tune individual models accordingly
- Want alerts when users tell personal stories, especially when they are about your brand? - Calibrate to your brand's terms
Our baseline ratios allow you to start reviewing content instantly, but all models are completely tunable to your specific brand's needs.
What are Machine Learning Models?
Machine Learning Models are programs that can find patterns or make decisions from a previously unseen dataset.
Machine Learning Models used in moderation are trained on very large data sets, usually vetted by experienced moderators who have rated the program’s decisions repeatedly, and used that info to retrain the model until the model is able to give a reliable rating on its own.
Integrations
Cleanspeak was built with the ability to fold in various Machine Learning Models into our interface, to allow our customers the ability to use more information to determine whether a piece of content is appropriate for their audience or not.
We are currently utilizing Google Perspective and SightEngine. These partnerships allow us to offer a suite of machine learning models, in multiple languages, integrated to seamlessly help give you more information to make decisions - without having to navigate those relationships on your own or integrate different models sets separately.
What are Machine Learning Models?
Machine Learning Models are programs that can find patterns or make decisions from a previously unseen dataset.
Machine Learning Models used in moderation are trained on very large data sets, usually vetted by experienced moderators who have rated the program’s decisions repeatedly, and used that info to retrain the model until the model is able to give a reliable rating on its own.