AI-powered incident categorization
Starting in 9.2, use artificial intelligence (AI) to automatically categorize incidents based on the incident summary and details in the Tickets application. The multi-class classifier (MCC) model provides suggestions for incident type and incident category fields to help you classify incidents more consistently and efficiently.
The AI-powered incident categorization feature uses machine learning to analyze the text in the Summary and Details fields of an incident record. Based on historical incident data, the MCC model suggests appropriate values for the Incident Type and Incident Category fields.
Benefits of AI-powered categorization
- Reduces manual effort in categorizing incidents that occur frequently but have minimal impact
- Improves consistency in incident classification across your organization
- Helps identify trends and patterns in incident data
- Enables better analysis of incident types and categories for preventive action
- Supports regulatory compliance by helping ensure that incidents are properly categorized
Prerequisites for using AI suggestions
Before users can access AI suggestions, an administrator must complete the following configuration tasks:
- Ensure that the AI service is available and running
- Configure the MCC model in the AI configuration application
- Define training and inference filters
- Select sufficient historical incidents for training
- Train the model and verify model accuracy
- Enable the incident type and incident category fields in the Tickets application user interface
For detailed configuration instructions, contact your system administrator.
How AI suggestions work
The MCC model analyzes the content of the Summary or Details fields, or both, to provide categorization suggestions. The model is trained on your organization's historical incident data to learn patterns and relationships between incident descriptions and their categories.
- The Summary field, the Details field, or both fields contain text
- The Incident Type field, the Incident Category field, or both fields are blank
If the Incident Type field is blank and the Incident Category field is populated, suggestions are displayed only for the Incident Type field. If the Incident Category field is blank and the Incident Type field is populated, suggestions are displayed only for the Incident Category field.
When suggestions appear
AI suggestions are displayed only when specific conditions are met:
- The incident must be included in the inference filter configured by your administrator
- The Summary field, Details field, or both fields contain text
- The Incident Type field, Incident Category field, or both fields are blank
- The model has been successfully trained and is ready for inferencing
If suggestions do not appear when expected, verify that these conditions are met or contact your system administrator.
Manual selection required
The MCC model does not apply suggestions automatically. Review the suggestions and manually select the one that best describes the incident. This approach helps ensure that you maintain control over the categorization process and can apply your expertise to each incident.