Create a Classification using ML Auto Classifier

Skyhigh provides ML Auto Classifiers to automatically detect and classify sensitive files based on Skyhigh pretrained Artificial Intelligence (AI) and Machine Learning (ML) models. You can use Skyhigh predefined ML auto classifiers to identify various types of sensitive files such as financial reports and statements, patient records, patents, source code, and ID files in sanctioned or shadow/web services. For details, see ML Auto Classifiers.

The ML Auto Classifier classification method allows you to define the ML Auto Classifiers, which can be used to match, process, and protect sensitive files during the DLP policy evaluation. This AI and ML-based data classification method helps to identify, and manage sensitive files and protect sensitive data leaving your organization.

You can use the ML Auto Classifier condition to define various types of ML Auto Classifiers based on Skyhigh predefined ML Auto Classifier categories with "is", "is one of", "is all of", "is not", or "is none of" operators. These operators precisely identify specific files by matching ML Auto Classifiers to trigger a match. This condition enables you to define DLP policies based on the ML Auto Classifiers identified in files to enforce tighter DLP policy controls on sensitive files. Using this automated data classification method, you can easily ensure the security of your sensitive data such as PII, financial records, healthcare records, and intellectual property. 

For example, you can create a classification using the ML Auto Classifier condition to automatically discover and categorize various types of financial files.

To create a classification using ML Auto Classifier:

1. Log in to Skyhigh CASB.
2. Go to Policy > DLP Policies > Classifications.
3. Under Actions, click Create Classification. 
   • Classification Name. Enter a name for this classification. For example, enter Financial Classification. Enter an optional description to describe its use or purpose.
     
   • Category. Select a Category from the list. For this example, select Finance and Trade.
   • Conditions. For IF, select ML Auto Classifier. The ML Auto Classifier cloud card appears.
     
4. Select an ML Auto Classifier from the list of supported ML Auto Classifiers and click Done. For example, select Finance and Trade (including Financial Reports and Financial Statements).
   • Finance and Trade
     • Financial Reports
     • Financial Statements
   • Healthcare
     • Patient Records
   • Intellectual Property
     • Patents
     • Source Code
   • PII
     • ID Documents (Image)
5. Operator. Select any one of the following operators to match files with specific, all, or none of the defined ML Auto Classifiers. By default, is or is not is selected.
   • is. Matches files if the selected ML-Auto Classifier is triggered.
   • is one of. Matches files if any one of the selected ML-Auto Classifiers is triggered.
   • is all of. Matches files if all of the selected ML-Auto Classifiers are triggered.
   • is not. Matches files if the selected ML-Auto Classifiers is not triggered.
   • is none of. Matches files if none of the selected ML-Auto Classifiers are triggered.
6. Click Save.

Your new Classification is saved to the selected category in the Classifications page. Add the classification to your DLP policies as required.​​​​​​