Toxicity and PII Filtering on Collaborative Mapping Submissions

To protect resident privacy and maintain community guidelines, all open-ended submissions on Collaborative Mapping activities are automatically screened for toxic language and personal information. This article explains how the filtering works.

What Is Filtered

The system automatically blocks submissions on open-ended Text questions and on 'Other' free-text fields if they contain:

• Toxic language that exceeds the platform toxicity threshold
• Phone numbers
• Email addresses
• Physical addresses
• URLs and links

How It Works

1. A resident enters text in an open-ended question or in an 'Other' field.
2. When the resident attempts to submit, the system screens the text using Google Perspective API and a personal information check.
3. If the text is acceptable, the submission proceeds.
4. If the text contains toxic language or personal information, the resident sees a message advising them that their submission does not meet community guidelines and asking them to revise it.
5. The resident must edit the text before they can submit.

Note: There is no moderation queue for Collaborative Mapping. Submissions that do not meet the criteria are blocked at the point of submission. Admins do not need to review individual submissions.