Edwin AI List-based Correlations for Models
Last updated - 22 August, 2025
Edwin AI models have the option to use list-based correlations, enabling you to group deduplicated alerts based on shared values in list-type fields, such as services, locations, or tags. List-based correlation is ideal when multiple alerts involve entities that support overlapping applications or features (for example, the same business service or infrastructure tier).
Models that use list-based correlation work by doing the following:
- Edwin evaluates alerts that contain values in a specified list-type field.
- Edwin calculates the number of exact matches between list items in the specified field.
- If the number of matching items equals or exceeds theminimum overlap count defined in the model, the alerts are grouped into a cluster.
Important: Only exact matches between list items are counted.
Configuring a list-based correlation requires setting the following parameters:
- Delimiter (for example, comma, semicolon)
- Trim rules (defining the clean brackets or quotes)
- Escape character (for special parsing)
For more information on how to configure a model using list-based correlations, see Edwin AI Model Creation.
List-based Correlation Scenarios
Use the following table to help you understand how Edwin applies list-based correlation logic:
Alerts:
- Alert A:
[inventory, cart, payment] - Alert B:
[inventory, orders]
| Scenario | Minimum Overlap Count | Overlapping Items | Result | Reason |
| 1 | 1 | [inventory] | Clustered | At least 1 matching item is found. |
| 2 | 2 | [inventory] | Not clustered | Only one match; required minimum is 2. |
| 3 | 3 | [inventory] | Not clustered | Only one match; required minimum is 3. |