แก้ไข

แชร์ผ่าน

Language role-based access control

  • บทความ

Azure AI Language supports Azure role-based access control (Azure RBAC), an authorization system for managing individual access to Azure resources. Using Azure RBAC, you assign different team members different levels of permissions for your projects authoring resources. See the Azure RBAC documentation for more information.

Enable Microsoft Entra authentication

To use Azure RBAC, you must enable Microsoft Entra authentication. You can create a new resource with a custom subdomain or create a custom subdomain for your existing resource.

Add role assignment to Language resource

Azure RBAC can be assigned to a Language resource. To grant access to an Azure resource, you add a role assignment.

  1. In the Azure portal, select All services.

  2. Select Azure AI services, and navigate to your specific Language resource.

    Note

    You can also set up Azure RBAC for whole resource groups, subscriptions, or management groups. Do this by selecting the desired scope level and then navigating to the desired item. For example, selecting Resource groups and then navigating to a specific resource group.

  3. Select Access control (IAM) on the left navigation pane.

  4. Select Add, then select Add role assignment.

  5. On the Role tab on the next screen, select a role you want to add.

  6. On the Members tab, select a user, group, service principal, or managed identity.

  7. On the Review + assign tab, select Review + assign to assign the role.

Within a few minutes, the target will be assigned the selected role at the selected scope. For help with these steps, see Assign Azure roles using the Azure portal.

Language role types

Use the following table to determine access needs for your Language projects.

These custom roles only apply to Language resources.

Note

  • All prebuilt capabilities are accessible to all roles
  • Owner and Contributor roles take priority over the custom language roles
  • Microsoft Entra ID is only used in case of custom Language roles
  • If you are assigned as a Contributor on Azure, your role will be shown as Owner in Language studio portal.

Cognitive Services Language Reader

A user that should only be validating and reviewing the Language apps, typically a tester to ensure the application is performing well before deploying the project. They may want to review the application’s assets to notify the app developers of any changes that need to be made, but do not have direct access to make them. Readers will have access to view the evaluation results.

Capabilities

API Access

  • Read
  • Test

All GET APIs under:

Cognitive Services Language Writer

A user that is responsible for building and modifying an application, as a collaborator in a larger team. The collaborator can modify the Language apps in any way, train those changes, and validate/test those changes in the portal. However, this user shouldn’t have access to deploying this application to the runtime, as they may accidentally reflect their changes in production. They also shouldn’t be able to delete the application or alter its prediction resources and endpoint settings (assigning or unassigning prediction resources, making the endpoint public). This restricts this role from altering an application currently being used in production. They may also create new applications under this resource, but with the restrictions mentioned.

Capabilities

API Access

  • All functionalities under Cognitive Services Language Reader.
  • Ability to:
  • Train
  • Write

Cognitive Services Language Owner

Note

If you are assigned as an Owner and Language Owner you will be shown as Cognitive Services Language Owner in Language studio portal.

These users are the gatekeepers for the Language applications in production environments. They should have full access to any of the underlying functions and thus can view everything in the application and have direct access to edit any changes for both authoring and runtime environments

Functionality

API Access

  • All functionalities under Cognitive Services Language Writer
  • Deploy
  • Delete

All APIs available under: