RelevanceAnalysis

{
  "relevant": boolean,
  "confidence": enum (ConfidenceLevel),
  "reasoning": string,
  "evidence": {
    object (Evidence)
  },
  "relevanceLevel": enum (RelevanceLevel)
}

• relevant: boolean — Indicates whether the threat is considered relevant.
• confidence: enum (ConfidenceLevel) — The level of confidence in the given verdict.
• reasoning: string — Human-readable explanation from the matcher, detailing why a particular result is considered relevant or not relevant.
• evidence: object (Evidence) — Evidence supporting the verdict, including matched and unmatched items.
• relevanceLevel: enum (RelevanceLevel) — The level of relevance.

Evidence

Details the evidence used to determine the relevance verdict.

{
  "commonThemes": [
    string
  ],
  "distinctThemes": [
    string
  ]
}

• commonThemes[]: string — A list of semantic themes or concepts found to be common, related, or aligned between the sources, supporting the verdict.
• distinctThemes[]: string — A list of semantic themes or descriptions unique to one source or semantically distant.

RelevanceLevel

Relevance Level of an issue.

• RELEVANCE_LEVEL_UNSPECIFIED: Default value, should never be set (source).
• RELEVANCE_LEVEL_LOW: Low Relevance.
• RELEVANCE_LEVEL_MEDIUM: Medium Relevance.
• RELEVANCE_LEVEL_HIGH: High Relevance.