When true, exportClassificationsInfer results collapse classificationCode/classificationDescription/classificationFullText to a binary "real uncontrolled code (e.g. EAR99)" vs "CONTROLLED" indicator instead of the exact ECCN.
The minimum probability mass the uncontrolled code's own prediction must have to be trusted. If the uncontrolled code's probability falls below this value, the best controlled candidate is selected instead, regardless of that candidate's own probability. Null disables this override.
The minimum probability mass at which the munitions tier claims an item, overriding a higher-scoring dual-use candidate. Deliberately low: this screen is recall-biased, so it flags for human review rather than asserting a classification. Null uses the service default.
When true, exportClassificationsInfer scores the jurisdiction's munitions list (US: the USML) ahead of its dual-use list and populates controlList, controlListVersion and controlTier. A munitions result is a review trigger, not a determination. Defaults to false for backward compatibility.
Fields
binaryDisplayEnabled
Boolean!
Required
exportClassificationsInferresults collapseclassificationCode/classificationDescription/classificationFullTextto a binary "real uncontrolled code (e.g. EAR99)" vs "CONTROLLED" indicator instead of the exact ECCN.confidenceThreshold
Decimal
createdAt
DateTime!
Required
createdBy
ID!
Required
id
ID!
Required
munitionsFlagThreshold
Decimal
munitionsScreeningEnabled
Boolean!
Required
exportClassificationsInferscores the jurisdiction's munitions list (US: the USML) ahead of its dual-use list and populatescontrolList,controlListVersionandcontrolTier. A munitions result is a review trigger, not a determination. Defaults to false for backward compatibility.organizationId
ID!
Required
updatedAt
DateTime!
Required
updatedBy
ID!
Required