Performs named-entity recognition on the given document

Notes:
Valid JSON cannot contain newline characters. These have to be escaped. (See also Interpretor documentation)
Fields text and url are mutually exclusive.
Examples:

{"text": "Hello world!"}