Yes, Text Moderation can be customized.
When using Ruled-based Text Moderationou can create multiple Custom text lists to apply different moderation criteria to different types of text items and force our API detect any words or content you feel should be flagged.
The entries of a list can be in the language you want and you can choose either to detect the exact string with no modifications using the exact match option or to also detect the possible variations and obfuscations by using the standard match option.
If you are using our Text Classification models, the API returns a score between 0 and 1 for each available class, reflecting how likely it is that someone would find the text problematic. Using the returned scores, you can determine thresholds for each class depending on how you want to moderate your content.
If you are using our Rule-based algorithms, the API returns intensity levels for each detected profanity. These levels can also be used as thresholds for your moderation.
Was this page helpful?