On March 29, 2018 we will be launching a host of updates to your Sift console. If you use Sift for any form of analysis of user-generated content, please read through the rest of this note. Also, if you are a currently signed up for our Content Abuse product, we encourage you to upgrade your Sift integration to get the full benefit of this launch. Please let us know if that’s something you’d like to learn more about.
Brand new Text Analysis Machine Learning models
We are launching machine learning models which identify risky patterns via deep analysis of user-generated text. These brand new models will be blended with our behavior learning models to get a complete understanding of the user to help you stop abuse before any objectionable content goes live on your site!
Specific Machine Learning features across different types of content
We are expanding our capability to differentiate between types of content, and assess risk across them.
- Listings
- Comments
- Profiles
- Messages
- Posts
- Reviews
This release comes with machine learning features tailored to catch bad behavior associated with the different types of content. For example, we will now analyze messaging activity like number of unique users contacted, and combine that with the profile information of those users.
Make Content Moderation a breeze with upgraded Review Queues
We’re introducing review queues which are focused on the content. The content to be reviewed will be front and center. We’ll also show you the risk signals to help make the content review even easier.
Flexibility to block specific content, in addition to the user
Apply Decisions on the Content: In addition to the existing capability of detecting and blocking fraudulent entities, we are introducing the ability to review specific content. For example, now you can block a specific comment while allowing the user to continue transacting on your site.
Create Content focused Workflows: Our Workflows automation platform will now allow you to create content-focused workflows. This means that you can define policies around each content creation event and take action on it. For example, now you can choose to filter out content which has certain words, and is created by new users.
Note that particularly long or sizable content sent to Sift may result in a 413 response code. If you encounter this code, feel free to reach out to support@sift.com and our team will take a closer look.