You can use the RedactFromText action to submit text in the Text HTTP parameter, and return the input text with any matches that contain confidential or sensitive data redacted.


http://localhost:7000/?action=RedactFromText&Text=The driver Joe Bloggs was questioned.

This action might result in output similar to the following XML:

<autn:redacted_output>The driver [redacted] was questioned</autn:redacted_output>


Parameter Description Required
AllowMultipleResults Whether to return multiple results for a single match. No
AllowOverlaps Whether to return more than one entity from any one section of text. No
CaseNormalization The case conversion to use for all incoming text. No
CJKNormalization How to normalize Chinese, Japanese, and Korean data before extraction. No
EnableComponents Whether to return all the components of the entity as well as the entity itself. No
EnableUniqueMatches Whether to return only unique matches in each document. No
Entities The entities to use for the extraction. No
Grammars The grammar files to use for the extraction. No
Locale Enables tokenization of Chinese, Japanese, Korean, and Thai languages. No
MatchCase Whether to ignore case when matching characters. No
MatchWholeWord Whether to take account of word boundaries when matching. No
MaxEntityLength The maximum number of characters in a returned entry. No
MinScore Matches only items with scores equal to or exceeding the threshold. No
NonGreedyMatch Whether to return the shortest match. No
RedactionOutputString A string that replaces redacted information in the output text. No
RedactionReplacementCharacter A single character that replaces each character in redacted text. No
TangibleCharacters A list of punctuation characters to treat as part of the word. No
Text The input text to process. Yes
TokenWithPunctuation Whether to treat all punctuation characters as part of a word token, rather than treating them as word boundaries No