AutomatedReasoningPolicyFidelityReport
A comprehensive analysis report that measures how accurately a generated policy represents the source documents. The report includes coverage and accuracy scores, detailed grounding information linking policy elements to source statements, and annotated document content.
Contents
- accuracyScore
-
A score from 0.0 to 1.0 indicating how accurate the policy rules are relative to the source documents. A higher score means the policy rules more faithfully represent the source material.
Type: Double
Valid Range: Minimum value of 0.0. Maximum value of 1.0.
Required: Yes
- coverageScore
-
A score from 0.0 to 1.0 indicating how well the policy covers the statements in the source documents. A higher score means more of the source content is represented in the policy.
Type: Double
Valid Range: Minimum value of 0.0. Maximum value of 1.0.
Required: Yes
- documentSources
-
A list of source documents with their content broken down into atomic statements and annotated with line numbers for precise referencing.
Type: Array of AutomatedReasoningPolicyReportSourceDocument objects
Required: Yes
- ruleReports
-
A mapping from rule identifiers to detailed fidelity reports for each rule, showing which source statements ground each rule and how accurate it is.
Type: String to AutomatedReasoningPolicyRuleReport object map
Key Length Constraints: Fixed length of 12.
Key Pattern:
[A-Z][0-9A-Z]{11}Required: Yes
- variableReports
-
A mapping from variable names to detailed fidelity reports for each variable, showing which source statements ground each variable and how accurate it is.
Type: String to AutomatedReasoningPolicyVariableReport object map
Key Length Constraints: Minimum length of 1. Maximum length of 64.
Key Pattern:
[A-Za-z][A-Za-z0-9_]*Required: Yes
See Also
For more information about using this API in one of the language-specific AWS SDKs, see the following: