Predefined Classifiers

Barracuda Data Inspector includes predefined (built-in) classifiers that automatically detect sensitive data across your Microsoft 365 environment (OneDrive and SharePoint). These classifiers cover personal identification, financial data, medical information, credentials, passports, driver's licenses, and tax identifiers across 30+ countries and regions.

In addition to predefined classifiers, you can create custom classifiers using keywords, keyword files, or regular expressions to detect organization-specific data types. See Custom Classifiers for instructions on creating custom classifiers.

How Predefined Classifiers Work

Predefined classifiers use multiple detection techniques to identify sensitive data with high accuracy and low false positives:

Pattern matching – Recognizes known formats for identifiers such as credit card numbers, Social Security numbers, passport numbers, and IBANs using structural patterns (e.g., a US SSN follows a 3-2-4 digit format with specific range constraints).
Checksum and validation algorithms – Applies mathematical validation where applicable. For example, credit card numbers are validated using the Luhn algorithm, IBANs use mod-97 check digits, and many national IDs include check-digit verification.
Contextual analysis – Examines surrounding text for contextual clues that increase detection confidence. A 9-digit number near the words "Social Security" scores higher than the same number in isolation.
Format normalization – Detects identifiers regardless of formatting variations (with or without dashes, spaces, or other separators).

These techniques work together – a detection typically requires both a structural pattern match and contextual confirmation before it is reported, reducing false positives while maintaining high recall.

OCR support: Predefined classifiers also apply to text extracted from images, screenshots, and scanned documents via Data Inspector's built-in OCR engine.

About Detection Patterns

The specific regular expressions, keyword dictionaries, and machine learning models that power predefined classifiers are proprietary and not published. This is an industry-standard practice – exposing exact detection patterns would allow bad actors to format sensitive data specifically to evade detection.

If you need to detect a specific data format that predefined classifiers do not cover, you can create a custom regex classifier with your own pattern.

Predefined Classifier Inventory

The following tables list every predefined classifier available in Barracuda Data Inspector, organized by category.

Credit Card

Classifier	Description

Classifier	Description
Credit Card Number	Detects major credit and debit card numbers (Visa, Mastercard, Amex, Discover, JCB, Diners Club, UnionPay, etc.) including Luhn check-digit validation.

Credentials and Secrets

Classifier	Description

Classifier	Description
Auth Token	Generic authentication tokens (Bearer tokens, OAuth tokens, session tokens).
Azure Auth Token	Microsoft Azure authentication tokens and service credentials.
Basic Auth Header	HTTP Basic Authentication headers containing encoded credentials.
Encryption Key	Cryptographic keys (AES, RSA, and other encryption key material).
GCP API Key	Google Cloud Platform API keys.
GCP Credentials	Google Cloud Platform service account credentials and key files.
Password	Plaintext passwords, including those in configuration files, connection strings, and code.
Weak Password Hash	Unsalted or weakly hashed passwords (MD5, SHA-1 without salt).
XSRF Token	Cross-site request forgery tokens that could be exploited if exposed.

Financial – Account Numbers

Classifier	Description

Classifier	Description
IBAN Code	International Bank Account Numbers used across 70+ countries, with mod-97 validation.
Canada Bank Account	Canadian bank account numbers (institution, transit, and account number format).
Japan Bank Account	Japanese bank account numbers (bank code, branch code, account number).

Financial — Identification

Classifier	Description

Classifier	Description
SWIFT Code	SWIFT/BIC codes used for international bank transfers (8 or 11 character format).
US Bank Routing (MICR)	US bank routing numbers as printed on checks in Magnetic Ink Character Recognition (MICR) format.
MICR Code	Bank check Machine Readable Zone codes (E-13B and CMC-7 formats).
American Bankers CUSIP ID	Committee on Uniform Securities Identification Procedures (CUSIP) identifiers for financial instruments.

Medical Codes

Classifier	Description

Classifier	Description
ICD-9 Code	International Classification of Diseases, 9th Revision diagnostic and procedure codes.
ICD-10 Code	International Classification of Diseases, 10th Revision diagnostic and procedure codes.
Medical Term	Medical terminology including drug names, conditions, procedures, and anatomical terms.

Medical Identification

Classifier	Description

Classifier	Description
US DEA Number	US Drug Enforcement Administration registration numbers assigned to healthcare providers.
US Healthcare NPI	US National Provider Identifier – 10-digit identification number for healthcare providers.

Personal Medical Identification

Classifier	Description

Classifier	Description
Australia Medicare Number	Australian Medicare card numbers (10-11 digits with check digit).
Canada British Columbia PHN	British Columbia Personal Health Number.
Canada OHIP	Ontario Health Insurance Plan number.
Canada Quebec HIN	Quebec Health Insurance Number (Régie de l'assurance maladie du Québec).
Scotland Community Health Index Number	Scottish CHI number used in NHS Scotland.
UK National Health Service Number	NHS number – 10-digit identifier for patients in England, Wales, and Isle of Man.

Passports

Classifier	Description

Classifier	Description
Passport (Generic)	Passport numbers from any country, including Machine Readable Zone (MRZ) detection from the passport's biometric data page.
US Passport	United States passport numbers (9 digits, alphanumeric).
Australia Passport	Australian passport numbers.
Canada Passport	Canadian passport numbers.
China Passport	People's Republic of China passport numbers.
France Passport	French passport numbers.
Germany Passport	German Reisepass numbers.
Japan Passport	Japanese passport numbers.
Korea Passport	Republic of Korea passport numbers.
Mexico Passport	Mexican passport numbers.
Netherlands Passport	Dutch passport numbers.
Poland Passport	Polish passport numbers.
Singapore Passport	Singaporean passport numbers.
Spain Passport	Spanish passport numbers.
Sweden Passport	Swedish passport numbers.
Taiwan Passport	Taiwan (ROC) passport numbers.
UK Passport	United Kingdom passport numbers.

Driver's Licenses

Classifier	Description

Classifier	Description
US Driver's License Number	United States driver's license numbers (format varies by state).
Australia Driver's License Number	Australian driver's license numbers (format varies by state/territory).
Canada Driver's License Number	Canadian driver's license numbers (format varies by province).
Germany Driver's License Number	German Führerschein numbers.
Japan Driver's License Number	Japanese driver's license numbers (12 digits).
Spain Driver's License Number	Spanish driver's license numbers.
UK Driver's License Number	United Kingdom driving licence numbers (DVLA format – name-encoded, 16 characters).

Personal Identification – National IDs

Classifier	Description

Classifier	Description
United States
US Social Security Number	9-digit SSN in XXX-XX-XXXX format, with area/group number validation.
US Adoption Taxpayer Identification Number (ATIN)	IRS-issued temporary number for children in domestic adoption.
Argentina
Argentina DNI Number	Documento Nacional de Identidad – 7-8 digit national identity number.
Belgium
Belgium National ID Card Number	Belgian eID card number.
Brazil
Brazil CPF Number	Cadastro de Pessoas Físicas – 11-digit individual taxpayer ID with check digits.
Canada
Canada Social Insurance Number	9-digit SIN with Luhn check digit.
Chile
Chile CDI Number	Cédula de Identidad – Chilean national identity number.
China
China Resident ID Number	18-digit PRC Resident Identity Card number (includes birth date and region encoding).
Colombia
Colombia CDC Number	Cédula de Ciudadanía – Colombian national identity number.
Denmark
Denmark CPR Number	Central Person Register number (personnummer) – 10 digits encoding date of birth.
Finland
Finland National ID Number	Finnish personal identity code (henkilötunnus).
France
France CNI	Carte Nationale d'Identité number.
France NIR	Numéro d'Inscription au Répertoire (Social Security number) – 13 digits + 2-digit key.
France Tax Identification Number	French tax ID (numéro fiscal).
Germany
Germany Identity Card Number	Personalausweis number.
Germany Taxpayer Identification Number	Steuerliche Identifikationsnummer (Steuer-IdNr) – 11 digits.
Germany SCHUFA ID	Credit bureau (SCHUFA) identification number.
Hong Kong
Hong Kong ID Number	HKID card number (letter prefix + 6 digits + check digit).
India
India Aadhaar Number	12-digit unique identity number issued by UIDAI, with Verhoeff check digit.
India PAN Number	Permanent Account Number – 10-character alphanumeric tax ID.
Indonesia
Indonesia NIK Number	Nomor Induk Kependudukan – 16-digit national identity number.
Italy
Italy Fiscal Code	Codice Fiscale – 16-character alphanumeric tax code encoding name, birth date, and location.
Japan
Japan Individual Number	My Number – 12-digit individual identification number.
Korea
Korea RRN	Resident Registration Number – 13 digits encoding date of birth and gender.
Mexico
Mexico CURP Number	Clave Única de Registro de Población – 18-character population registry code.
Netherlands
Netherlands BSN Number	Burgerservicenummer – 9-digit citizen service number with 11-test check.
Norway
Norway National ID Number	Fødselsnummer – 11-digit personal identification number.
Paraguay
Paraguay CIC Number	Cédula de Identidad Civil.
Peru
Peru DNI Number	Documento Nacional de Identidad – 8-digit national ID.
Poland
Poland PESEL Number	11-digit Universal Electronic System for Registration of the Population number.
Poland National ID Number	Polish ID card number.
Portugal
Portugal CDC Number	Cartão de Cidadão – Citizen Card number.
Singapore
Singapore NRIC Number	National Registration Identity Card number (letter + 7 digits + check letter).
Spain
Spain CIF Number	Certificado de Identificación Fiscal – tax identification for legal entities.
Spain DNI Number	Documento Nacional de Identidad – 8 digits + check letter.
Spain NIE Number	Número de Identidad de Extranjero – foreigner identification number.
Spain NIF Number	Número de Identificación Fiscal – tax identification number.
Spain Social Security Number	Spanish Social Security affiliation number.
Sweden
Sweden National ID Number	Personnummer – 10 or 12 digits encoding date of birth.
Thailand
Thailand National ID Number	13-digit Thai national identification number.
Turkey
Turkey ID Number	TC Kimlik Numarası – 11-digit Turkish identification number.
United Kingdom
UK National Insurance Number	NINO – 2 letters + 6 digits + 1 letter (e.g., AB123456C).
UK Taxpayer Reference	Unique Taxpayer Reference (UTR) – 10-digit HMRC identifier.
Uruguay
Uruguay CDI Number	Cédula de Identidad – Uruguayan identity document number.
Venezuela
Venezuela CDI Number	Cédula de Identidad – Venezuelan identity document number.

Tax Identification

Classifier	Description

Classifier	Description
US Individual Taxpayer Identification Number (ITIN)	IRS-issued 9-digit number for individuals not eligible for an SSN (9XX range).
US Preparer Taxpayer Identification Number (PTIN)	IRS-issued number for paid tax return preparers.
Australia Tax File Number	8-9 digit Australian tax identifier with check digit.
India GST Number (Individual)	Goods and Services Tax Identification Number for individuals.

Personal Identifiable Information (PII) – General

Classifier	Description

Classifier	Description
Date of Birth	Dates of birth in various formats (MM/DD/YYYY, DD-Mon-YYYY, etc.) when accompanied by contextual indicators.
Gender	Gender identifiers when part of personal records or forms.
Email Address	Email addresses in standard format (local-part@domain).

Location

Classifier	Description

Classifier	Description
Street Address	Physical street addresses including street number, name, city, state/province, and postal code.

Other Sensitive Identifiers

Classifier	Description

Classifier	Description
IMEI Hardware ID	International Mobile Equipment Identity – 15-digit device identifier for mobile phones.
FDA Code	US Food and Drug Administration product or registration codes.
US Employer Identification Number (EIN)	IRS-issued 9-digit number for businesses (XX-XXXXXXX format).
US Vehicle Identification Number (VIN)	17-character vehicle identifier with check-digit validation.
Credit Card Track Number	Magnetic stripe track data from credit/debit cards (Track 1 and Track 2).

Geographic Coverage Summary

Data Inspector's predefined classifiers cover identifiers from the following regions:

Region	Countries/Areas

Region	Countries/Areas
Global / Generic	Credit cards, IBAN, SWIFT, passports (generic), credentials, medical codes, PII, email
United States	SSN, driver's license, passport, EIN, ITIN, PTIN, DEA, NPI, ATIN, VIN, and more
Europe	Germany, France, Spain, UK, Netherlands, Sweden, Finland, Norway, Denmark, Poland, Portugal, Italy, Belgium, Scotland
Asia-Pacific	Australia, Japan, Korea, India, China, Singapore, Thailand, Indonesia, Hong Kong, Taiwan
Americas (non-US)	Canada, Brazil, Argentina, Chile, Colombia, Mexico, Paraguay, Peru, Uruguay, Venezuela
Middle East / Africa	Turkey

Classifier Categories in the Product UI

UI Category	What It Covers	Label Color

UI Category	What It Covers	Label Color
Credit Card	Credit and debit card numbers	Blue
Credentials	Passwords, API keys, auth tokens, encryption keys	Blue
Passport	Passport numbers from 17 countries + generic/MRZ	Blue
License Number	Driver's license numbers from 7 countries	Blue
Personal ID	National identification numbers from 30+ countries	Blue
Personal Medical ID	Health insurance and medical ID numbers (AU, CA, UK)	Blue
Tax ID	Taxpayer identification numbers (US, AU, IN)	Blue
Financial ID	Bank accounts, routing numbers, SWIFT/BIC, CUSIP	Blue
Suspicious	Sharing violations and potentially suspicious content patterns	Yellow
Malicious	Malware, viruses, and other malicious payloads (via ATP)	Red
Custom	Custom keyword, keyword file, and regex classifier matches	Gray

On the Detections page, each flagged file displays classification labels indicating what was found. The number shown inside a label is the count of times that classifier matched within the file.

Confidence and Sensitivity

Each detection includes a confidence level that reflects how certain Data Inspector is that the identified data is a true match. Higher confidence means more validation signals were present (structural match, checksum validation, and contextual confirmation).

Not all classifiers have the same baseline sensitivity. Some identifiers (like credit card numbers with Luhn validation) can be detected with very high confidence from structure alone, while others (like street addresses) rely more heavily on contextual signals and may require surrounding indicators to confirm a detection.

Compliance Framework Mapping

Predefined classifiers support detection requirements for common compliance frameworks:

Framework	Key Classifiers

Framework	Key Classifiers
PCI DSS	Credit Card Number, Credit Card Track Number
HIPAA	Medical codes (ICD-9, ICD-10), Medical Terms, US DEA Number, US Healthcare NPI, Personal Medical IDs
GDPR	National IDs (EU countries), Date of Birth, Email Address, Street Address
CCPA	US SSN, Driver's License, Financial identifiers, Email, Date of Birth
SOX	Financial identifiers, Credentials, EIN
FERPA	Student records containing PII (SSN, Date of Birth, Email Address, Street Address) – combine with custom classifiers for student IDs
GLBA	Financial identifiers, SSN, Credit Card, Bank Account numbers

For compliance frameworks that require detection of organization-specific identifiers (like student IDs for FERPA, or employee IDs for internal policies), use custom classifiers to supplement the predefined set.

Frequently Asked Questions

Can I see the exact patterns used by predefined classifiers? No. The detection patterns and algorithms used by predefined classifiers are proprietary and not published. This is an industry-standard security practice – if exact patterns were public, it would be straightforward to format sensitive data to evade detection. If you need to detect a specific format, you can create a custom regex classifier with your own pattern.

Can I disable specific predefined classifiers? Predefined classifiers cannot be individually disabled through the product UI. All predefined classifiers are active by default and will generate detections when matches are found. To manage unwanted detections, you have two options:

Create a policy and use the "Ignore classifier" action – This suppresses detections for the specified classifier from appearing on the Detections page, reducing noise from classifiers that are not relevant to your organization. The Scan log will still capture the scanning history for that file. The file continues to be evaluated by other policies. See Policies for details.
Use the "Dismiss" remediation action – This moves individual detections from the Active tab to the Resolved tab, clearing them from your active view without modifying the file. Dismissed detections remain in the Scan Log for audit purposes.

Can I request a new predefined classifier? Yes. Contact Barracuda Support or your account team with the data type you need detected. Include the format specification and any validation rules if known. For immediate needs, a custom regex classifier can typically be created to cover the gap while a predefined classifier is evaluated.

Do predefined classifiers work on images and scanned documents? Yes. Data Inspector's built-in OCR engine extracts text from images (JPG, PNG), scanned PDFs, and screenshots. The same predefined classifiers are then applied to the extracted text. OCR-based detection is automatic – no additional configuration is required.

How are false positives handled? Data Inspector includes an optional feedback mechanism that allows admins to report false positives and false negatives directly from the file preview. This feature is not enabled by default – contact Barracuda Support to activate it. Because the feedback workflow requires removing redactions to reveal the underlying sensitive data, enabling it means any DI user with admin rights can view unredacted content during review. For the full step-by-step workflow, see Provide Feedback to Barracuda. To suppress detections from a specific classifier entirely, create a policy with the "Ignore classifier" action – this prevents those detections from appearing on the Detections page while the file continues to be evaluated by remaining policies. To clear individual false positives from your active view, use the "Dismiss" action on the Detections page.