Validation of library documents is the application of a generic OK marker that can be interpreted in various ways depending on the users' work style.
For example, in projects that produce models trained with a library, the validation of training library documents may indicate that all expected results—be they categories or extractions—have been correctly annotated. If you work in a team, typically there are users who annotate and others who validate their annotations.
Validating a document in a test library can have the same meaning: the expected results have all been correctly annotated.
By filtering the library documents based on the value of the validation flag, users can easily find the documents that have not yet been validated and work only on those.
In categorization and extraction projects, when launching an experiment, it is possible to indicate one's preference with respect to the validated documents of the training library, that is one can decide to consider only those documents or prefer them in the case of sub-sampling.
Validating and invalidating is possible in detail view—and its variants: immersive view, PDF document view, condensed view—and in context view.
To validate a document select Validate this document , to remove validation select Remove validation form this document .
The validation flag is visible in all document lists.