Skip to content

Latest commit

 

History

History
83 lines (39 loc) · 4.23 KB

working-with-intelligent-recommendations-340651c.md

File metadata and controls

83 lines (39 loc) · 4.23 KB

Working with Intelligent Recommendations

You can create intelligent recommendation model to manage recommedations for the approval tasks.

Prerequisites

Context

Intelligent recommendations assist users in making decision to approve or reject specific tasks. These recommendations are based on the confidence level generated for the task. Confidence level is a numeric value calculated by a trained AI model. This value assists the user in deciding the action to take on a specific task. Confidence levels and intelligent recommendations improve productivity for end users who need to approve or reject a task. Users can see the confidence level for the assigned task to make an informed decision.

To create an intelligent recommendation model, follow the procedure:

Procedure

  1. In the Task Center Administration tile, navigate to the Intelligent Recommendations tab.

  2. Select Create Model option to create a new intelligent recommendation model.

  3. In the Create Intelligent Recommendation Model dialog box, provide the following details:

    • Task Provider: Select the required provider or connector from the dropdown list. The dropdown only displays the task providers that support intelligent recommendations.

    • Task Type: Select the task for which you want to configure a recommendation.

    • Model Name: Enter a name for the model.

    • Description: Enter a brief description about the model.

    Configure the following Task Type Attributes:

    • Numerical Attributes: Select the attributes from the dropdown list whose values might be subject to change. For example: NetValue, RequiredQuantity.

    • Categorical Attributes: Select the attributes from the dropdown list whose values are definite and may not change. For example: Product ID, Product Name, SalesOrganization.

    Note:

    To create a model, you must configure at least five Numerical and Categorical attribute values together.

    Initially, the Categorical Attributes and the Numerical Attributes dropdowns display all the custom attributes. However, when a custom attribute is selected in one category, it's no longer displayed in the other category.

  4. Choose Create.

    An intelligent recommendation model is created and is listed under Intelligent Recommendations with the Preparing Data status. Once the data is ready, the model is set to training with the In Training status.

  5. Once the model is Ready for Activation, select Activate option.

    The status of the model is set to Active and the intelligence recommendation capabilities are accessible to the users.

    Tip:

    • To edit a model, choose the model from the list and choose ✏️.
    • To deactivate a model, choose the required model from the list and choose Deactivate option.
    • To delete a model, choose the required model from the list and choose:wastebasket:.

Related Information

Working with the Task List

Data Protection and Privacy