Create RegEx Classifications
Objective: Create RegEx-Based Data Classification
In this section, we are going to build some data identifiers from scratch using regular expressions. These expressions will form the universal building blocks for a basic DLP policy that you could deploy using most DLP tools, including those built into many IaaS, PaaS, and SaaS platforms.
Don’t worry, if you’re not an expert with regular expressions (and so few are), you can use Skyhigh’s AI-assisted regular expression builder!
Tasks
Create a Regular Expression Classification for French Social Security Numbers
-
Access your data classifications. From the Policy heading, select DLP Policies and then Classifications.
-
Note the long list of pre-canned data classifications that are maintained by Skyhigh. We’re going to ignore these for the moment and build our classifications from scratch instead.
-
From the Action drop-down menu, select Create Classification.
-
Provide a name for the classification, such as “French SSN”.
-
In the Conditions section, under Rule group 1, click Select Criteria and select Advanced Pattern from the list that appears.
-
Click the New button to create a new Advanced Pattern. You can give this pattern the same name (French SSN).
-
Use the AI RegEx Generator by clicking the button and then telling the generative AI that you would like a regular expression to describe “French Social Security numbers”. You can either cut and paste the pattern that appears or use the Insert RegEx button to insert it automatically (if it appears).
-
Click Save to save the pattern, Done, and then Save to save the classification.
Data classifications can use multiple regular expressions, boolean logic, dictionaries (lists), and proximity, which we will cover more closely later in the workshop.
Repeat the Process for Credit Card Numbers and UK Drivers Licenses
Use the same procedure as above to create two more regular expression-based classifications for credit card numbers and UK driver’s licenses. When finished, you should have three custom classifications: