DLP inference example notebook

models/validation-inference-scripts/dlp-models/data/config.json

@@ -0,0 +1,63 @@
+{
+    "entity_labels": [
+        "medical_record_number",
+        "date_of_birth",
+        "ssn",
+        "date",
+        "first_name",
+        "email",
+        "last_name",
+        "customer_id",
+        "employee_id",
+        "name",
+        "street_address",
+        "phone_number",
+        "ipv4",
+        "credit_card_number",
+        "license_plate",
+        "address",
+        "user_name",
+        "device_identifier",
+        "bank_routing_number",
+        "date_time",
+        "company_name",
+        "unique_identifier",
+        "biometric_identifier",
+        "account_number",
+        "city",
+        "certificate_license_number",
+        "time",
+        "postcode",
+        "vehicle_identifier",
+        "coordinate",
+        "country",
+        "api_key",
+        "ipv6",
+        "password",
+        "health_plan_beneficiary_number",
+        "national_id",
+        "tax_id",
+        "url",
+        "state",
+        "swift_bic",
+        "cvv",
+        "pin"
+    ],
+    "type_weights": {
+        "password": 85,
+        "credit_card": 90,
+        "ssn": 95,
+        "address": 60,
+        "email": 40,
+        "phone_us": 45,
+        "phone_numbers": 45,
+        "ip_address": 30,
+        "date": 20,
+        "api_key": 80,
+        "customer_id": 65,
+        "personal": 70,
+        "financial": 85,
+        "health": 75,
+        "api_credentials": 75
+    }
+}

models/validation-inference-scripts/dlp-models/data/regex_patterns.json

@@ -0,0 +1,307 @@
+{
+    "medical_record_number": [
+        "\\b(?:MRN|MR)[-_#:]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b",
+        "\\bmedical[-_]?record[-_]?number[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b",
+        "\\b\\d{5,12}\\b"
+    ],
+    "date_of_birth": [
+        "\\b(?:dob|date[_\\s]?of[_\\s]?birth|birth[_\\s]?date)[:=]?\\s*\\d{1,2}[\\/\\-\\.]\\d{1,2}[\\/\\-\\.]\\d{2,4}\\b",
+        "\\b(?:dob|birth[_\\s]?date)[:=]?\\s*\\d{4}[\\/\\-\\.]\\d{2}[\\/\\-\\.]\\d{2}\\b",
+        "\\b(?:dob|birth[_\\s]?date)[:=]?\\s*\\w{3,9}\\s\\d{1,2},\\s\\d{4}\\b",
+        "\\b(?:DOB)\\s*[:=]?\\s*\\d{2}\\s*\\w{3,9}\\s*\\d{4}\\b",
+        "\\b(?:birthdate|bday)[:=]?\\s*\\d{2}-\\w{3}-\\d{4}\\b"
+    ],
+    "ssn": [
+        "\\b\\d{3}[\\s|-]\\d{2}[\\s|-]\\d{4}\\b",
+        "\\b\\d{9}\\b"
+    ],
+    "date": [
+        "\\b\\d{1,2}[\\/\\-\\.]\\d{1,2}[\\/\\-\\.]\\d{2,4}\\b",
+        "\\b\\d{4}[\\/\\-\\.]\\d{2}[\\/\\-\\.]\\d{2}\\b",
+        "\\b\\w{3,9}\\s\\d{1,2},\\s\\d{4}\\b",
+        "\\b\\d{1,2}\\s\\w{3,9}\\s\\d{4}\\b",
+        "\\b\\w{3,9}\\s\\d{4}\\b"
+    ],
+    "first_name": [
+        "\\b(?:first[_\\s]?name|First|FName|Given[_\\s]?name|Forename)[:=]?\\s*[A-Z][a-z]+\\b",
+        "\\b[A-Z][a-z]{1,20}[-'][A-Z][a-z]{1,20}\\b",
+        "\\b(?:first name|fname)[:=]\\s*['\"]?([A-Za-z-']{2,25})['\"]?\\b"
+    ],
+    "email": [
+        "\\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Za-z]{2,}\\b",
+        "\\bemail[:=]?\\s*[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Za-z]{2,}\\b",
+        "\\bE-mail[:=]?\\s*[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Za-z]{2,}\\b",
+        "\\bEmailAddress[:=]?\\s*[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Za-z]{2,}\\b",
+        "\\b[A-Za-z0-9._%+-]+\\s?\\(at\\)\\s?[A-Za-z0-9.-]+\\.[A-Za-z]{2,}\\b"
+    ],
+    "last_name": [
+        "\\b(?:last[_\\s]?name|Last|Surname|LName|Family[_\\s]?name)[:=]?\\s*[A-Z][a-z]+\\b",
+        "\\b[A-Z][a-z]{1,20}[-'][A-Z][a-z]{1,20}\\b",
+        "\\b(?:last name|lname)[:=]\\s*['\"]?([A-Za-z-']{2,25})['\"]?\\b"
+    ],
+    "customer_id": [
+        "\\bcustomer[_\\s]?id[:=]?\\s*[A-Z0-9]{6,12}\\b",
+        "\\b(CustID|Client[_\\s]?ID|CID|CustomerNumber)[:=]?\\s*\\w{6,12}\\b",
+        "\\b(?:CUST|CUS|C)[-_#]\\d{3,10}\\b",
+        "\\bcustomer[-_]?id[:=]\\s*['\"]?([A-Za-z0-9-_]{3,15})['\"]?\\b",
+        "\\b[A-Z]{1,3}\\d{4,10}\\b"
+    ],
+    "employee_id": [
+        "\\b(?:employee[_\\s]?id|EmpID|Staff[_\\s]?Number|EmployeeNumber|WorkerID)[:=]?\\s*\\w{4,12}\\b",
+        "\\b(?:E|EMP|EMPLOYEE)[-_#]\\d{3,10}\\b",
+        "\\b\\d{6,10}\\b",
+        "\\b[A-Z]\\d{5,8}\\b",
+        "\\bemployee[-_]?id[:=]\\s*['\"]?([A-Za-z0-9-_]{3,15})['\"]?\\b"
+    ],
+    "name": [
+        "\\b(?:name|Full[_\\s]?Name|Person|Contact|Client)[:=]?\\s*[A-Z][a-z]+(?:\\s[A-Z][a-z]+)+\\b",
+        "\\b[A-Z][a-z]{1,20}\\s+[A-Z][a-z]{1,20}\\b",
+        "\\b[A-Z][a-z]{1,20}\\s+[A-Z]\\.\\s+[A-Z][a-z]{1,20}\\b",
+        "\\bname[:=]\\s*['\"]?([A-Za-z \\.-]{2,50})['\"]?\\b",
+        "\\b[A-Z][a-z]{1,20}\\s+[A-Z][a-z]{1,20}[-][A-Z][a-z]{1,20}\\b"
+    ],
+    "phone_number": [
+        "\\(\\d{3}\\)\\s?\\d{3}-\\d{4}",
+        "\\b\\d{3}-\\d{3}-\\d{4}\\b",
+        "\\b\\d{10}\\b",
+        "\\b\\+\\d{1,2}\\s?\\(?\\d{3}\\)?[\\s.-]\\d{3}[\\s.-]\\d{4}\\b",
+        "\\bPhone[:=]?\\s*\\d{3}[\\s.-]?\\d{3}[\\s.-]?\\d{4}\\b",
+        "\\b\\d{3}[-. ]?\\d{3}[-. ]?\\d{4}\\b",
+        "\\b\\(\\d{3}\\)[-. ]?\\d{3}[-. ]?\\d{4}\\b",
+        "\\b\\+?\\d{1,3}[-. ]?\\d{3}[-. ]?\\d{3}[-. ]?\\d{4}\\b",
+        "\\b\\d{5}[-. ]?\\d{5}\\b"
+    ],
+    "ipv4": [
+        "\\b(?:\\d{1,3}\\.){3}\\d{1,3}\\b",
+        "\\bIPv4[:=]?\\s*(?:\\d{1,3}\\.){3}\\d{1,3}\\b",
+        "\\b\\d{1,3}\\s*\\.\\s*\\d{1,3}\\s*\\.\\s*\\d{1,3}\\s*\\.\\s*\\d{1,3}\\b",
+        "\\b(?:25[0-5]|2[0-4]\\d|[01]?\\d?\\d)(?:\\.(?:25[0-5]|2[0-4]\\d|[01]?\\d?\\d)){3}\\b",
+        "\\bIP[:=]?\\s*(?:\\d{1,3}\\.){3}\\d{1,3}\\b"
+    ],
+    "credit_card_number": [
+        "\\b(?:\\d[ -]*?){13,16}\\b",
+        "\\b\\d{4}[ -]?\\d{4}[ -]?\\d{4}[ -]?\\d{4}\\b",
+        "\\b(?:Visa|MasterCard|Amex|Discover)[:=]?\\s*\\d{4}[ -]?\\d{4}[ -]?\\d{4}[ -]?\\d{4}\\b",
+        "\\bCardNumber[:=]?\\s*\\d{13,16}\\b",
+        "\\b(?:\\d{4}\\s){3}\\d{4}\\b"
+    ],
+    "license_plate": [
+        "\\b[A-Z]{1,3}-\\d{1,4}[A-Z]{0,2}\\b",
+        "\\b\\d{1,4}[A-Z]{1,3}\\b",
+        "\\b[A-Z]{1,7}\\d{1,4}\\b",
+        "\\bPlate[:=]?\\s*[A-Z0-9-]{4,10}\\b",
+        "\\b[A-Z]{1,3}[-\\s]?\\d{1,4}[-\\s]?[A-Z]{0,3}\\b",
+        "\\b\\d{1,3}[-\\s]?[A-Z]{1,3}[-\\s]?\\d{1,4}\\b",
+        "\\blicense[-_]?plate[:=]\\s*['\"]?([A-Za-z0-9-_\\s]{2,10})['\"]?\\b"
+    ],
+    "address": [
+        "\\b\\d{1,5}\\s+[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\s+(Street|St|Avenue|Ave|Road|Rd|Lane|Ln|Drive|Dr)\\b",
+        "\\bAddress[:=]?\\s*\\d{1,5}\\s+[A-Za-z\\s]+\\s+(Street|St|Avenue|Ave|Rd|Dr|Blvd)\\b",
+        "\\b\\d+\\s+[A-Z][a-z]+\\s+(St|Ave|Blvd|Rd)\\b",
+        "\\b\\d+\\s+[A-Za-z\\s]+\\s+(Street|Avenue|Lane|Road|Drive)\\s*(Apt\\.?\\s*\\d+)?\\b",
+        "\\b\\d+\\s+[A-Za-z0-9\\s]+,\\s*[A-Za-z\\s]+,\\s*[A-Z]{2}\\s*\\d{5}(-\\d{4})?\\b",
+        "\\baddress[:=]\\s*['\"]?([A-Za-z0-9 \\.,#-]{10,100})['\"]?\\b",
+        "\\b\\d{1,5}\\s+[A-Za-z0-9 \\.-]{5,50},?\\s+[A-Za-z ]{2,30},?\\s+[A-Za-z ]{2,20}\\s+\\d{5}(?:-\\d{4})?\\b"
+    ],
+    "user_name": [
+        "\\busername[:=]?\\s*[A-Za-z0-9._-]{4,}\\b",
+        "\\buser[:=]?\\s*[A-Za-z0-9._-]{4,}\\b",
+        "\\buname[:=]?\\s*[A-Za-z0-9._-]{4,}\\b",
+        "\\bLogin[:=]?\\s*[A-Za-z0-9._-]{4,}\\b",
+        "\\baccount[:=]?\\s*[A-Za-z0-9._-]{4,}\\b",
+        "\\buser[-_]?name[:=]\\s*['\"]?([A-Za-z0-9_-]{3,20})['\"]?\\b",
+        "\\b@[a-zA-Z0-9_-]{3,20}\\b"
+    ],
+    "device_identifier": [
+        "\\bDevice[-_\\s]?ID[:=]?\\s*[A-Fa-f0-9:\\-]{10,}\\b",
+        "\\bIMEI[:=]?\\s*\\d{14,17}\\b",
+        "\\bSerial[:=]?\\s*[A-Z0-9]{8,}\\b",
+        "\\bDevice[:=]?\\s*[A-Z0-9]{6,}\\b",
+        "\\bMAC[:=]?\\s*(?:[A-Fa-f0-9]{2}[:\\-]){5}[A-Fa-f0-9]{2}\\b"
+    ],
+    "bank_routing_number": [
+        "\\bRouting[-_\\s]?Number[:=]?\\s*\\d{9}\\b",
+        "\\bABA[:=]?\\s*\\d{9}\\b",
+        "\\bRouting[:=]?\\s*\\d{9}\\b",
+        "\\bBankRoutingID[:=]?\\s*\\d{9}\\b",
+        "\\b\\d{9}\\s*(?:Routing)?\\s*Number\\b"
+    ],
+    "date_time": [
+        "\\b\\d{4}-\\d{2}-\\d{2}[ T]\\d{2}:\\d{2}:\\d{2}\\b",
+        "\\b\\d{2}/\\d{2}/\\d{4}\\s+\\d{1,2}:\\d{2}(?:\\s*[APap][Mm])?\\b",
+        "\\b\\w{3,9}\\s\\d{1,2},\\s\\d{4}\\s+at\\s+\\d{1,2}:\\d{2}\\b",
+        "\\b\\d{1,2}:\\d{2}\\s*[APap][Mm],?\\s*\\d{1,2}[\\/\\-]\\d{1,2}[\\/\\-]\\d{2,4}\\b",
+        "\\b\\d{2}:\\d{2}:\\d{2}\\s+\\d{4}-\\d{2}-\\d{2}\\b"
+    ],
+    "company_name": [
+        "\\b(?:Inc|LLC|Corp|Ltd|Limited|Corporation|Company)\\b",
+        "\\bcompany[-_]?name[:=]\\s*['\"]?([A-Za-z0-9 \\.-]{2,50})['\"]?\\b"
+    ],
+    "unique_identifier": [
+        "\\bUID[:=]?\\s*[A-Za-z0-9\\-]{8,}\\b",
+        "\\bID[:=]?\\s*[A-Za-z0-9]{8,}\\b",
+        "\\bUnique[-_ ]?ID[:=]?\\s*[A-Za-z0-9\\-]{6,}\\b",
+        "\\bUUID[:=]?\\s*[A-Fa-f0-9\\-]{36}\\b",
+        "\\b[A-Fa-f0-9]{8}\\-[A-Fa-f0-9]{4}\\-[A-Fa-f0-9]{4}\\-[A-Fa-f0-9]{4}\\-[A-Fa-f0-9]{12}\\b"
+    ],
+    "biometric_identifier": [
+        "\\b(?:Fingerprint|IrisScan|FaceID|BiometricID|VoicePrint)[:=]?\\s*[A-Za-z0-9+/=]{20,}\\b",
+        "\\b(?:fingerprint|retina|iris|face|voice)[-_]?id[:=]\\s*['\"]?([A-Za-z0-9-_]{8,30})['\"]?\\b",
+        "\\b[A-F0-9]{8}-[A-F0-9]{4}-[A-F0-9]{4}-[A-F0-9]{4}-[A-F0-9]{12}\\b",
+        "\\bbiometric[-_]?data[:=]\\s*['\"]?([A-Za-z0-9-_]{8,30})['\"]?\\b"
+    ],
+    "account_number": [
+        "\\bAccount\\s*Number[:=]?\\s*\\d{6,12}\\b",
+        "\\bAcct\\.?[:=]?\\s*\\d{6,12}\\b",
+        "\\bAcc#[:=]?\\s*\\d{6,12}\\b",
+        "\\bAccount[:=]?\\s*\\d{6,12}\\b",
+        "\\b\\d{6,12}\\b\\s*Account\\b",
+        "\\b\\d{8,17}\\b",
+        "\\b\\d{4}[-\\s]\\d{4}[-\\s]\\d{4}[-\\s]\\d{4}\\b",
+        "\\baccount[-_]?number[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b",
+        "\\bacc[-_]?no[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b"
+    ],
+    "city": [
+        "\\bCity[:=]?\\s*[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bLocated in\\s+[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bLives in\\s+[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bResides in\\s+[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bcity[:=]\\s*['\"]?([A-Za-z \\.-]{2,30})['\"]?\\b"
+    ],
+    "certificate_license_number": [
+        "\\bLicense\\s*No\\.?[:=]?\\s*[A-Za-z0-9\\-]{6,}\\b",
+        "\\bCert(?:ificate)?\\s*ID[:=]?\\s*[A-Za-z0-9\\-]{6,}\\b",
+        "\\bLic#[:=]?\\s*[A-Za-z0-9]{6,}\\b",
+        "\\bReg\\.?\\s*No[:=]?\\s*[A-Za-z0-9]{6,}\\b",
+        "\\bPermit\\s*Number[:=]?\\s*[A-Za-z0-9]{6,}\\b",
+        "\\b(?:Certificate|License)\\s+(?:Number|No|#)[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b",
+        "\\b[A-Z]{1,3}-\\d{5,10}\\b",
+        "\\b\\d{2}-\\d{5,8}\\b",
+        "\\b[A-Z]\\d{6,10}[A-Z]\\b"
+    ],
+    "time": [
+        "\\b\\d{1,2}:\\d{2}\\s*[APap][Mm]\\b",
+        "\\b\\d{2}:\\d{2}:\\d{2}\\b",
+        "\\bTime[:=]?\\s*\\d{1,2}:\\d{2}\\b",
+        "\\b\\d{1,2}\\s*(AM|PM)\\b",
+        "\\b\\d{2}h\\d{2}\\b"
+    ],
+    "postcode": [
+        "\\b\\d{5}(?:-\\d{4})?\\b",
+        "\\bZip[:=]?\\s*\\d{5}(-\\d{4})?\\b",
+        "\\bPostal\\s*Code[:=]?\\s*\\d{5}(?:-\\d{4})?\\b",
+        "\\b\\d{3}\\s?\\d{2}\\b",
+        "\\b\\d{4,6}\\b"
+    ],
+    "vehicle_identifier": [
+        "\\bVIN[:=]?\\s*[A-HJ-NPR-Z0-9]{17}\\b",
+        "\\bVehicle\\s*ID[:=]?\\s*[A-HJ-NPR-Z0-9]{17}\\b",
+        "\\bChassis\\s*Number[:=]?\\s*[A-HJ-NPR-Z0-9]{17}\\b",
+        "\\bEngine\\s*No\\.?[:=]?\\s*[A-Za-z0-9]{6,}\\b",
+        "\\bV\\.?I\\.?N\\.?[:=]?\\s*[A-HJ-NPR-Z0-9]{17}\\b",
+        "\\b[A-HJ-NPR-Z0-9]{17}\\b",
+        "\\bVIN[:=]\\s*['\"]?([A-HJ-NPR-Z0-9]{17})['\"]?\\b",
+        "\\bvehicle[-_]?id[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b"
+    ],
+    "coordinate": [
+        "\\b[-+]?\\d{1,3}\\.\\d+,\\s*[-+]?\\d{1,3}\\.\\d+\\b",
+        "\\bLat[:=]?\\s*-?\\d{1,2}\\.\\d+\\s*Long[:=]?\\s*-?\\d{1,3}\\.\\d+\\b",
+        "\\bLatitude[:=]?\\s*-?\\d{1,2}\\.\\d+\\s*Longitude[:=]?\\s*-?\\d{1,3}\\.\\d+\\b",
+        "\\b\\(\\s*-?\\d{1,3}\\.\\d+\\s*,\\s*-?\\d{1,3}\\.\\d+\\s*\\)\\b",
+        "\\bGPS\\s*Coord(?:inate)?s?[:=]?\\s*-?\\d{1,3}\\.\\d+\\s*,\\s*-?\\d{1,3}\\.\\d+\\b"
+    ],
+    "country": [
+        "\\bCountry[:=]?\\s*[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bNationality[:=]?\\s*[A-Z][a-z]+\\b",
+        "\\bFrom\\s+[A-Z][a-z]+(?:\\s[A-Z][a-z]+)*\\b",
+        "\\bCitizenship[:=]?\\s*[A-Z][a-z]+\\b",
+        "\\bResident of\\s+[A-Z][a-z]+\\b",
+        "\\b[A-Z]{2,3}\\b",
+        "\\bcountry[:=]\\s*['\"]?([A-Za-z ]{2,56})['\"]?\\b"
+    ],
+    "api_key": [
+        "\\b(?:Key|Token|Secret|Auth)[:=]?\\s*[A-Za-z0-9\\-_]{20,}\\b",
+        "\\b(?:[A-Za-z0-9+/]{40,})\\b",
+        "\\b[A-Za-z0-9]{32,}\\b",
+        "\\b(?:key|api[-_]?key|access[-_]?token)[-_]?['\"]?([A-Za-z0-9_\\-\\.]{16,})['\"]?\\b",
+        "\\b[a-zA-Z0-9_]{8,}\\.[a-zA-Z0-9_]{8,}\\.[a-zA-Z0-9_-]{8,}\\b"
+    ],
+    "ipv6": [
+        "(?:[0-9a-fA-F]{1,4}:){7}[0-9a-fA-F]{1,4}",
+        "((?:[0-9A-Fa-f]{1,4}(?::[0-9A-Fa-f]{1,4})*)?)::((?:[0-9A-Fa-f]{1,4}(?::[0-9A-Fa-f]{1,4})*)?)",
+        "::1",
+        "\\bIPv6[:=]?\\s*['\"]?([0-9a-fA-F]{1,4}:){7}[0-9a-fA-F]{1,4}['\"]?\\b"
+    ],
+    "password": [
+        "\\bPassword[:=]?\\s*.+\\b",
+        "\\bPwd[:=]?\\s*[^\\s]{6,}\\b",
+        "\\bPass[:=]?\\s*[^\\s]{6,}\\b",
+        "\\bUserPass[:=]?\\s*[^\\s]{6,}\\b",
+        "\\bSecret[:=]?\\s*[^\\s]{6,}\\b",
+        "\\b(?:password|pwd|pass)[:=]\\s*['\"]?([^\\s\"']{6,30})['\"]?\\b"
+    ],
+    "health_plan_beneficiary_number": [
+        "\\bHPBN[:=]?\\s*\\d{9,12}\\b",
+        "\\bBeneficiary\\s*ID[:=]?\\s*\\d{9,12}\\b",
+        "\\bHealth\\s*Plan\\s*No\\.?[:=]?\\s*\\d{9,12}\\b",
+        "\\bPlanID[:=]?\\s*\\d{9,12}\\b",
+        "\\bInsurance[:=]?\\s*\\d{9,12}\\b",
+        "\\b[A-Z]\\d{8}\\b",
+        "\\b\\d{3}-\\d{2}-\\d{4}\\b",
+        "\\bhealth[-_]?plan[-_]?(?:number|id|no)[:=]\\s*['\"]?([A-Za-z0-9-_]{5,20})['\"]?\\b",
+        "\\b[A-Z]{1,5}\\d{6,12}\\b"
+    ],
+    "national_id": [
+        "\\bNID[:=]?\\s*\\d{6,14}\\b",
+        "\\bNational\\s*ID[:=]?\\s*\\d{6,14}\\b",
+        "\\bNatID[:=]?\\s*\\d{6,14}\\b",
+        "\\bCitizen\\s*ID[:=]?\\s*\\d{6,14}\\b",
+        "\\bIdentity\\s*Number[:=]?\\s*\\d{6,14}\\b"
+    ],
+    "tax_id": [
+        "\\bTIN[:=]?\\s*\\d{9}\\b",
+        "\\bTax\\s*ID[:=]?\\s*\\d{9}\\b",
+        "\\bEIN[:=]?\\s*\\d{2}-\\d{7}\\b",
+        "\\bTaxpayer\\s*Number[:=]?\\s*\\d{9}\\b",
+        "\\b\\d{3}-\\d{2}-\\d{4}\\s*\\(Tax\\)\\b"
+    ],
+    "url": [
+        "\\bhttps?:\\/\\/[^\\s]+\\b",
+        "\\bwww\\.[A-Za-z0-9\\-\\.]+\\.[a-z]{2,}\\b",
+        "\\bURL[:=]?\\s*https?:\\/\\/[^\\s]+\\b",
+        "\\bLink[:=]?\\s*https?:\\/\\/[^\\s]+\\b",
+        "\\b(?:http|https):\\/\\/\\S+\\b"
+    ],
+    "state": [
+        "\\bState[:=]?\\s*[A-Z][a-z]+\\b",
+        "\\bLives in\\s+[A-Z][a-z]+\\b",
+        "\\bLocated in\\s+[A-Z][a-z]+\\b",
+        "\\b[A-Z]{2}\\s*State\\b",
+        "\\b\\([A-Z]{2}\\)\\b",
+        "\\bstate[:=]\\s*['\"]?([A-Za-z ]{2,20})['\"]?\\b"
+    ],
+    "swift_bic": [
+        "\\bSWIFT[:=]?\\s*[A-Z]{6}[A-Z0-9]{2}([A-Z0-9]{3})?\\b",
+        "\\bBIC[:=]?\\s*[A-Z]{6}[A-Z0-9]{2,5}\\b",
+        "\\bSWIFT-BIC[:=]?\\s*[A-Z0-9]{8,11}\\b",
+        "\\bSWIFT Code[:=]?\\s*[A-Z0-9]{8,11}\\b",
+        "\\b\\b[A-Z]{4}[A-Z]{2}[A-Z0-9]{2}([A-Z0-9]{3})?\\b"
+    ],
+    "cvv": [
+        "\\bCVV[:=]?\\s*\\d{3,4}\\b",
+        "\\bSecurity\\s*Code[:=]?\\s*\\d{3,4}\\b",
+        "\\bCVC[:=]?\\s*\\d{3,4}\\b",
+        "\\bCard\\s*Verification[:=]?\\s*\\d{3,4}\\b",
+        "\\b\\d{3,4}\\s*\\(CVV\\)\\b",
+        "\\b\\d{3,4}\\b",
+        "\\b(?:CVV|CVC|CVV2|CID)[:=]\\s*['\"]?(\\d{3,4})['\"]?\\b",
+        "\\bsecurity code[:=]\\s*['\"]?(\\d{3,4})['\"]?\\b"
+    ],
+    "pin": [
+        "\\bPIN[:=]?\\s*\\d{4,6}\\b",
+        "\\bPin\\s*Code[:=]?\\s*\\d{4,6}\\b",
+        "\\bAccess\\s*Code[:=]?\\s*\\d{4,6}\\b",
+        "\\b\\d{4,6}\\s*\\(PIN\\)\\b",
+        "\\bPasscode[:=]?\\s*\\d{4,6}\\b"
+    ]
+}