The purpose of this ontology is to annotate W3C PROV statements with degrees of uncertainty. It provides a simple way to annotate uncertain provenance and provenance of uncertainty.
AssertionType
Generic class to be used as value of the :assertionType property, to provide context for the uncertainty aspects of the provenance you are asserting. Use one of the predefined subclasses, or define your own.
Complete
Indicates that all information about this provenance statement is known (e.g., all members of a prov:Collection are asserted).
Future
Indicates that this provenance describes a process that is yet to be executed (i.e., so called 'prospective provenance').
HumanAsserted
Indicates that a provenance statement was generated by a human.
Incomplete
Indicates that not all information about this provenance statement is known (e.g., a prov:Collection could have more, unknown members other than those asserted).
MachineCollected
Indicates that a provenance statement was collected/observed by a machine.
MachineGenerated
Indicates that a provenance statement was automatically generated by a machine (e.g., using a heuristic).
Trusted
Indicates that the asserter deems this provenance 100% certain.
Untrusted
Indicates that the asserter has doubts about the trustworthiness of this provenance.
assertionConfidence
This property allows you to annotate an uncertain (qualified) provenance relation, and denotes the confidence assigned to a provenance statement by the asserter. For example, this could be used to annotate a statement that was automatically generated with less than 100% confidence. You should provide as much context about the confidence score as possible (e.g., the tool used, the meaning of the score, etc.) to make comparison with other values possible.
assertionType
This attribute describes the type of uncertainty associated with a provenance statement.
assertionConfidence
This attribute specifies a confidence score, denoting how confident a user or application was about the content whose provenance is asserted. You should provide as much context about the confidence score as possible (e.g., the tool used, the meaning of the score, etc.) to make comparison with other values possible.