SharePoint modern pages support approvals with preconfigured Microsoft Flow. What if we want to partially automate and improve this process with some smart services, which will automatically reject approval if a page contains not appropriate content. Or add a piece of information to an approver to pay special attention to a particular page? This is a nice place for Azure cognitive services to come into the play! 

For your convenience, the whole flow in png format is available by this link. 

Azure Content Moderator

Azure provides a lot of different cognitive services and one of them is Content Moderator:

The Azure Content Moderator API is a cognitive service that checks text, image, and video content for material that is potentially offensive, risky, or otherwise undesirable.

Sounds interesting! To be more precise, currently Content Moderator has below core features:

• Detects profanity in over 100 languages. It includes profane words based on a built-in list. You have an option to provide your own inclusion or exclusion lists with your specific terms. 
• Text classification. It detects a potential presence of language that may be considered sexually explicit or adult or offensive in certain situations. Text receives classification from 0 to 1. The higher the score, the higher the model is predicting that the category may be applicable (text is offensive or adult etc.)
• Detects Personally Identifiable Information (PII) - any presence of emails, SSN, emails, phones, addresses in text. This is sensitive information and sometimes it's not desirable to have it available publicly.
• Image and video moderation detects adult or racy content, face detection, and some other features

That's just basics. The content moderator is able to do a lot more cool things. More...