Skip to content

JSON for resources

The following describes the format of the JSON file that is produced when you export classes from an extraction project and that you can use during the creation wizard to define classes from a file.

{
    "classGroups": array,
    "classes": array
}

The classGroups array contains the definition of classes associated with groups, while the classes array contains the definition of classes not associated with any group.

The structure of the classGroups item is:

{
    "name": string
    "classes": array
}

where name (string) is the name of the group and the classes array contains the definition of the classes in the group.

The structure of the item of the classes array is the same as the items of the classes array at the higher level (classes not associated with any group):

{
    "name": string,
    "abbreviation_key": string,
    "colour": string,
    "extraction_flag": boolean,
    "is_favourite": boolean,
    "positional": boolean
}

where:

  • name is the name of the class.
  • abbreaviation_key is a three-character code used as an abbreviation for the class name. You cannot use the code of an entity type as an abbreviation.
  • colour is the hexadecimal number of a color in RGB notation to be used to graphically identify the class in the various visualizations.
  • extraction_flag indicates whether or not the occurrences of the class should be extracted by the project's predictive models.
  • is_favourite indicates whether the class is among the favorites. Multiple classes can be marked as favorites. The classes marked as favorites are shown first in the list during annotation.
  • positional is true if the class is positional, false is it is a metadata class.