1. Implementing Help Desk
  2. Configure Profile Options

Configure Profile Options

You can set options for classifying help desk requests by configuring some email profile options. These profile options are predefined, but you can change their values based on your requirements.

Here are the profile options to configure followed by the procedure for configuration:

Profile Options for Classifying HR Help Desk Requests

Profile Option Description
ORA_SVC_ENABLE_ATTR_DETECTION_MASTER_IDCS_APP Enables the use of the master IDCS application authentication for SR classification REST API calls.

The default value is No.

Caution:

You must enable this profile option by setting the value to Yes, otherwise Help Desk can’t connect to the AI application.
ORA_SVC_ENABLE_INBOUND_MSG_HRHD_CATEGORY_PREDICTION Enables the prediction for the category attribute during the processing of inbound emails to create help desk requests.

The default value is Yes.

To turn off category predictions, set the value to No.
ORA_SVC_ENABLE_INBOUND_MSG_HRHD_PRODUCT_PREDICTION Enables the prediction for the product attribute during the processing of inbound emails to create help desk requests.

The default value is Yes.

To turn off product predictions, set the value to No.
ORA_SVC_INBOUND_MSG_PREDICTION_THRESHOLD_FOR_HRHD_CATEGORY_ID Specifies the threshold value to decide whether the predicted category ID can be stamped on the help desk request.

The default value is: 0.6

This means that if the prediction has a confidence level of more than 60%, only then you stamp the predicted value on the help desk request. Otherwise, don’t stamp it on the request.

It’s possible to run the service in the dark mode by setting the threshold value to 1.0. In the dark mode, the prediction happens but it isn’t stamped on the request. This way, you can see what the correct labels would have been if the accuracy had been set lower.

ORA_SVC_INBOUND_MSG_PREDICTION_THRESHOLD_FOR_HRHD_PRODUCT_ID Specifies the threshold value to decide whether the predicted product ID can be stamped on the help desk request.

The default value is: 0.7

This means that if the prediction is more than 70% accurate, only then you stamp the predicted value on the request. Otherwise, don’t stamp it on the request.

It’s possible to run the service in the dark mode by setting the threshold value to 1.0. In the dark mode, the prediction happens but it isn’t stamped on the request. This way, you can see what the correct labels would have been if the accuracy had been set lower.

Profile Options for Internal Help Desk Requests

Profile Option Description
ORA_SVC_ENABLE_ATTR_DETECTION_MASTER_IDCS_APP Enables the use of the master IDCS application authentication for SR classification REST API calls.

The default value is No.

Caution:

You must enable this profile option by setting the value to Yes, otherwise Help Desk can’t connect to the AI application.
ORA_SVC_ENABLE_INBOUND_MSG_ISR_CATEGORY_PREDICTION Enables the prediction for the category attribute during the processing of inbound emails to create help desk requests.

The default value is Yes.

To turn off category predictions, set the value to No.
ORA_SVC_ENABLE_INBOUND_MSG_ISR_PRODUCT_PREDICTION Enables the prediction for the product attribute during the processing of inbound emails to create internal help desk requests.

The default value is Yes.

To turn off product predictions, set the value to No.
ORA_SVC_INBOUND_MSG_PREDICTION_THRESHOLD_FOR_ISR_CATEGORY_ID Specifies the threshold value to decide whether the predicted category ID can be stamped on the internal help desk request.

The default value is: 0.6

This means that if the prediction has a confidence level of more than 60%, only then you stamp the predicted value on the internal help desk request. Otherwise, don’t stamp it on the request.

It’s possible to run the service in the dark mode by setting the threshold value to 1.0. In the dark mode, the prediction happens but it isn’t stamped on the request. This way, you can see what the correct labels would have been if the accuracy had been set lower.

ORA_SVC_INBOUND_MSG_PREDICTION_THRESHOLD_FOR_ISR_PRODUCT_ID Specifies the threshold value to decide whether the predicted product ID can be stamped on the internal help desk request.

The default value is: 0.7

This means that if the prediction is more than 70% accurate, only then you stamp the predicted value on the request. Otherwise, don’t stamp it on the request.

It’s possible to run the service in the dark mode by setting the threshold value to 1.0. In the dark mode, the prediction happens but it isn’t stamped on the request. This way, you can see what the correct labels would have been if the accuracy had been set lower.
  1. Sign in as a setup user or administrator.
  2. Navigate to the Setup and Maintenance work area.
  3. Open the Tasks panel tab and click Search.
  4. Search for and click the Manage Internal Service Request Classification Profile Options or the Manage HR Service Request Classification Profile Options task.
  5. On the Manage Internal Service Request Classification Profile Options or the Manage HR Service Request Classification Profile Options page, click the profile options listed in the previous table.
  6. Do the following for each profile option:
    1. In the Profile Values region, in the Profile Level drop-down list, ensure that Site is selected.
    2. For the first three profile options, select Yes from the Profile Value drop-down list.
    3. To change the value for the last two profile options, specify a value in the Profile Value field as explained in the table.
  7. Click Save.