Instruction
This article introduces the implementation process of ZIM content moderation, including service activation and operation.
Service Activation
Content moderation service is free within 7 days of activation. Starting from the 8th day, the service will be billed based on usage. It is recommended that customers test ZIM before applying for content moderation service.
1. Create a NEXTDATA account
Contact the ZEGOCLOUD sales representative or email global_support@zegocloud.com with the information shown in the table below to create a NEXTDATA account (used for content moderation):
| Field | Description | Purpose |
|---|---|---|
| Company Name | Your company name。 | - |
| Account Email | It should be the same as the email address associated with the ZEGOCLOUD console account. | Used for logging into the NEXTDATA console and checking for duplicate accounts. |
| Industry | Your business industry. Available options include social networking, online education, E-commerce platform, and gaming. Default to social networking. | Used to configure the default moderation policy for NEXTDATA. |
2. Activate ZIM content moderation service and configure moderation policies
After creating the NEXTDATA account, provide the information shown in the table below to activate the content moderation service.
| Configuration | Required | Description |
|---|---|---|
| Review Level | Yes | The review level determines whether ZIM will continue sending a message when it receives a review result of "uncertain". The options include (choose one):
|
| Language | Yes | Supports the following languages (choose one):
|
| Text Moderation | Yes | Currently fixed to "Default Policy", which moderates content related to politics, terrorism, prohibition, pornography, abuse, advertisement, privacy, and advertising law violations. If adjustments are needed, please contact ZEGOCLOUD technical support. |
| Image Moderation | Yes | Currently fixed to "Default Policy", which moderates content related to politics, pornography, terrorism, QR code recognition, and advertising, and supports text violation detection in images. If adjustments are needed, please contact ZEGOCLOUD technical support. |
| Voice Moderation | No | Currently fixed to "Default Policy", which moderates content related to politics, pornography, terrorism, and supports statesman voiceprints, advertising, prohibition, and abuse detection. If adjustments are needed, please contact ZEGOCLOUD technical support. |
| Video Moderation | No | Video moderation includes moderation of both images and audio in videos. Currently fixed to "Default Policy". Images: Default moderation includes content related to politics, pornography, terrorism, and supports QR code recognition, and text violation detection in images. If adjustments are needed, please contact ZEGOCLOUD technical support. Audio: Default moderation includes content related to politics, pornography, terrorism, and supports statesman voiceprints, advertising, prohibition, and abuse detection. If adjustments are needed, please contact ZEGO technical support. |
Please select the appropriate moderation level and moderation policies based on your business needs.
Operation
After completing the service activation mentioned above, you can directly use the built-in content moderation service of ZIM.
When a user sends a message, ZIM will review the message based on the moderation policies, thresholds, and strengths set for the corresponding message channel (one-on-one chat, group chat, room), and perform relevant business operations based on the moderation result.
If a message fails to be sent due to moderation, you will receive error code 6000221. You can perform relevant business operations based on this error code.
If you need to adjust the moderation policies, please contact ZEGOCLOUD technical support. The keys for configuring moderation policies are as follows:
- Custom moderation policies for different message sources: Now, ZIM supports setting different moderation policies for different message channels. For example, if you use ZIM in one-on-one chat and group chat scenarios, you can set different moderation thresholds for the "one-on-one chat" and "group chat" to meet your needs. For more details, please refer to Overview - Built-in content moderation capability.
- Moderation level affects how ZIM handles messages based on moderation results: The moderation results include: "pass", "uncertain", or "fail".
- If you set the moderation level to Weak Moderation: ZIM will continue sending the message when the moderation result is "uncertain".
- If you set the moderation level to Strong Moderation: ZIM will stop sending the message when the moderation result is "uncertain". You can set the corresponding moderation level based on your business needs.
- Choose whether to send before moderation, or moderate before sending: When sending image messages in one-on-one chats and group chats, you can choose "send before moderation" or "moderate before sending" based on your needs to maximize the responsiveness of messages in your business. Similar restrictions include:
- Text messages only support "moderate before sending*".
- Due to the fact that ZIM does not store messages in rooms, in order to maximize the security and compliance of messages, text and image messages in rooms only support "moderate before sending".
- For audio and video messages, due to their large file sizes and special usage scenarios, there is a longer moderation delay. ZIM defaults to "moderate before sending" for these types of messages.