Configuring intelligent category prediction in incidents with BMC Helix classifier

Service desk managers can use AI to classify incoming incidents into appropriate categories. AI analyzes historical, closed incidents in BMC Helix ITSM to predict the appropriate categories. The service capability uses the Helix model to predict the categories.

Use this capability for the following benefits:

Better classification of incidents into categories, resulting in quicker resolutions.
Reduced effort to categorize incoming incidents.

Scenario to perform category prediction

Carl, a Service Desk Agent at Apex Global, is handling a high volume of incoming IT support tickets. He receives a ticket with a vague description and no categorization. He clicks under the categorization fields.
With the new capability, the system instantly analyzes the ticket content and suggests appropriate categories based on historical data. He reviews all suggestions, makes a minor adjustment, and submits the ticket.

Workflow for category prediction

Before you begin

Make sure that you have the Cognitive Service Config permission. For information about this permission, see Permissions-to-configure-the-cognitive-service.
Make sure that you understand the types of training data and the guidelines to follow while generating the data.
Obtain the URL and credentials required to configure the cognitive service.

To configure intelligent category prediction in incidents with BMC Helix classifier

To configure the CCS parameters

Log in to Mid Tier.
Select AR System Administration > AR System Administration Console > System > General > Centralized Configuration.
From the Component Name list, select com.bmc.arsys.server.shared > *.
Set the value of the following CCS parameters:
- Classification-Service-Provider CCS parameter as HELIX.
  Use this parameter to select the provider of the classification service.
- Cognitive-Service-Confidence-Threshold CCS parameter as 0.0.
Click Apply.

To configure the cognitive service connection settings

Log in to Mid Tier
Navigate to Application Administration Console > Custom Configuration > Foundation > Advanced Options > Cognitive Service Configuration.
On the Cognitive Data Setup form, click Cognitive Service Setup.
On the Cognitive Service Connection Settings form, enter the field values for User Name, Password, and Cognitive Service URL.
The URL and credentials are provided to the admin user.
In Enable Cognitive Service, select Yes.
The value that you select updates the Enable-Cognitive-Service CCS parameter.
After you save the settings, click Test Connection to check whether the cognitive service is configured on your system.
- If cognitive service is successfully configured, the following message is displayed:
  Connection Successful (ARNOTE 10000)
- If the system encounters problems while configuring the service, an error is displayed.
(Optional) If the system fails to configure the service due to an error, check the error details in the arextension.log file in the ARSysteminstalldirectory\Arserver\Db location and re-enter the correct values on the form.

To generate training data to train the algorithm on closed incidents of the past

Log in to Mid Tier.
Navigate to Application Administration Console > Custom Configuration > Foundation > Advanced Options > Cognitive Service Configuration.
From the Company list, select the company for which you want to enable the capability.
From the Application list, select Incident Management.
From the Template list, select one of the following values:
- Select Incident Categorization Training Data to train the data for the Operational category and Product category fields.
- Select Incident Resolution Categorization Training Data to train the data for the Resolution category and Resolution product category fields.
Click Add Selected Template.
Select the added template and click Generate Training Data.

The AI algorithm analyzes closed incidents from the selected company in BMC Helix ITSM, and considers them for model training.

If the training data generation fails, see Troubleshooting incident requests.

Recommendation

Each support group must have at least 100 incidents to train the model such that BMC Helix ITSM can predict categories of new incidents.
We recommend setting the Cognitive-Service-Confidence-Threshold CCS parameter as 0.0. This ensures that the Helix model can predict categories of new incidents.
To improve the overall incident management, we recommend that you train the models of the following cognitive service capabilities:
- Categorization prediction
- Support group prediction

Where to go from here

Creating an incident request