易盾

NetEase Yidun Text ReCognition is a content safety service designed to review user-generated text before it reaches public-facing platforms. It is part of the broader Yidun product family from NetEase, a Chinese Technology company, and targets teams that need automated screening of posts, comments, messages, and other text-heavy content. The service is presented as a cloud-based tool with an online trial area, so businesses can test bASIc detection behavior before committing to a more tAIlored setup.

What it does

The tool analyzes incoming text and flags content that may violate platform rules or APPlicable regulations. Rather than relying on a single keyword list, it combines pattern matching with models tuned for specific categories of unwanted text. The vendor describes Personalized matching models and customized detection schemes, which suggests configurations can be adjusted for different communities and risk profiles.

Key capabilities

• adult content screening: IDEntifies sexually explicit descriptions and related material.
• Advertising and promotion filtering: detects text intended to drive traffic, promote products, or divert users to other channels.
• Prohibited item detection: flags references to goods or activities restricted by law or policy.
• Sensitive and political content checks: looks for content that may affect public order or spread rumors, following the provider's stated policy categories.
• Violence and extremism recognition: identifies text describing violent, graphic, or abusive themes.
• Abuse and profanity detection: catches insults, personal attacks, and vulgar language.
• Low-value spam detection: recognizes repetitive or meaningless filler content.

Who it is for

This service suits social media platforms, forums, blogs, community apps, live chat products, and marketplaces that handle large volumes of user submissions. Trust and safety teams, moderators, and platform operators can use it as a first-pass filter so human reviewers focus on edge cases. It may also interest developers building moderation into comment systems or messAGIng features.

How to get started

Prospective users can visit the trial area, paste Sample text, and run detection to observe how the system categorizes results. The provider notes that the online trial only dEMOnstrates partial functionality. Teams with needs such as foreign-language detection or custom labels are advised to contACT the vendor for a one-to-one customized service. Exact onboarding steps, integration options, and response formats may vary, so confirm details with the provider.

Pricing and access caveat

Public Information does not specify Standard pricing tiers, quotas, or contract terms; costs may vary based on usage volume and customization. The trial should be treated as an evaluation tool rather than a full production test. Because detection policies reflect regional legal and cultural standards, organizations operating in multiple markets should verify that category DeFinitions align with their own moderation guidelines before deployment.