Scripts

Forcepoint DLP provides a list of built-in script classifiers. Many are written in Python, a development language that mimics natural language, and some are written in C++. (See NLP Scripts.)

Script classifiers are most often used to classify numeric data such as credit card numbers and Social Security numbers. Because the scripts are optimized for this purpose, script classifiers are more accurate than regular expression classifiers. Scripts analyze both content and context using statistical analysis or decision trees.

Note that fingerprinting is better than scripts at detecting the exact credit card numbers in your database—for example, your customers’ credit card numbers.

For catching credit card information in general, however, use the script classifier. Scripts detect any valid credit card number.

Fingerprinting and a script classifier may be used in combination with different levels of severity and different actions.

Scripts can also be used to classify software design documents, source code (C, C++, C# and Java), SPICE, Verilog (Verilog hardware design source code), and VHDL (VHDL and VHDL AMS hardware design source code).

To view a list of script content classifiers:

  1. Click Main > Policy Management > Content Classifiers.
  2. Select Patterns & Phrases.
  3. Filter the Classifier Type column to display only scripts.

Click Delete in the toolbar at the top of the content pane to delete a selected classifier, or click Where Used to see where the classifier is used.

The Used in a Policy column in the table indicates whether the classifier is used in a policy at all.

You cannot generate your own scripts, but you can edit an existing script, change its parameters, and save it under a new name.

Click a classifier name to view or edit properties. Add the classifier to a rule to activate it in a policy.

Forcepoint can create custom classifiers for a specific organization on request. Talk to your Sales representative for more details.