Zscaler provides the following Data Loss Prevention (DLP) dictionaries. Dictionaries marked with an asterisk (*) are not supported for Endpoint DLP. To learn more, see About Endpoint Data Loss Prevention (DLP). To learn more about configuring predefined DLP dictionaries, see Editing Predefined DLP Dictionaries.

• Aadhaar Card Number (India)

  This dictionary detects Aadhaar Unique Identification (UID) numbers from India.

  The popular format for an Aadhaar UID number is a 12-digit number, separated at the 4th and 8th digits by a delimiter. The delimiter can be a period, hyphen, or space.

  The following are examples of popular formats:

  • NNNNNNNNNNNN
  • NNNN-NNNN-NNNN

  This dictionary uses the Verhoeff checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Aadhaar UID number matches a valid range. The Aadhaar UID number can contain: A period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. An alphabetical boundary.	The number formats that can trigger the dictionary are: Q2161 6729 3627O 8384-2795-9970 A number format like 2@075-8515-612d5 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Aadhaar UID number is in a popular format. The Aadhaar UID number can contain a non-alphanumeric boundary. It cannot be bound by alphabetical characters. The Aadhaar UID number must use the same delimiters for the full number.	The number formats that can trigger the dictionary are: @216167293627@ 8384-2795-9970 The number formats that do not trigger the dictionary are: 2075-8515/6125 F838427959970B
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Aadhaar UID number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Aadhaar Card, UID, or UIDAI number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: @216167293627@ 8384-2795-9970 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 2075-8515/6125 F838427959970B

  Close
• ABA Bank Routing Numbers

  This dictionary detects ABA routing transit numbers from the United States.

  The popular format for an ABA routing transit number is a 9-digit number that has no delimiters.

  An example of a popular format is NNNNNNNNN.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the ABA routing transit number matches a valid range. The ABA routing transit number can contain a period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed.	The number formats that can trigger the dictionary are: 021---302---567 053 9021 97 QQ036001808// 011...600...033 The number formats that do not trigger the dictionary are: 50dd86d97067 aa8543210bb74
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The ABA routing transit number is in a popular format. The ABA routing transit number can not contain a period, hyphen, or space as delimiters. No delimiters are allowed.	The number formats that can trigger the dictionary are: 053902197 036001808 011600033 The number formats that do not trigger the dictionary are: 021---302---567 50dd86d97067 aa8543210bb74
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The ABA routing transit number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, aba routing number, aba number, american bank association routing number, bank routing number, or routing transit number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 021302567 053902197 036001808 011600033 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 50dd86d97067 aa8543210bb74

  Close
• Addresses (Japan)

  This dictionary is disabled by default. To access this feature, contact your Zscaler Account team.

  This dictionary detects content related to addresses from Japan.

  This dictionary does not use a checksum.

  The following are examples of acceptable formats for Japanese Addresses:

  • 〒064-0807 北海道札幌市中央区南七条西２ー４ー４ (full address with hyphen as delimiter)
  • 〒064-0807 北海道札幌市中央区南七条西２丁目４番４号 (full address with kanji]
  • 札幌市中央区南七条西2の4の4 (postal code optional)
  • 札幌市中央区南七条西2丁目4-4 (postal code is optional)
  • 新宿区高田馬場1-7-2 (postal code and prefecture are optional)
  • 新宿区高田馬場1丁目7の2 (postal code and prefecture are optional)

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Adult Content

  This dictionary detects adult or mature content.

  The popular format for adult or mature content is sexually explicit keywords, as well as foul language words.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Australia Passport Number

  This dictionary detects passport numbers (AUPP) from Australia.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the AUPP number matches a valid range.	A number format like N1234567 triggers the dictionary. The number formats that do not trigger the dictionary are: N 1234567 N12345 67 Q1234567 (invalid first letter)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The AUPP number is in a popular format.	A number format like N1234567 triggers the dictionary. The number formats that do not trigger the dictionary are: N 1234567 N12345 67
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The AUPP number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, passport, passport number, passportno, and passport num.	A number format like N1234567 triggers the dictionary. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: N 1234567 N12345 67

  Close
• Bulgaria Uniform Civil Number

  This dictionary detects uniform civil numbers (EGN) from Bulgaria.

  This dictionary uses the Modulo 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the EGN number matches a valid range.	The number formats that can trigger the dictionary are: 752 316 9263 7 -52/.316 926...- 3 A number format like 7523169264 (invalid checksum) will not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The EGN number is in a popular format.	A number format like 7523169263 triggers the dictionary. The number formats that do not trigger the dictionary are: 752 316 9263 7 -52/.316 926...- 3
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PTTIN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, bucn, egn, vim, uniform civil number, and unified civil number.	A number format like 7523169263 triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 752 316 9263 7 -52/.316 926...- 3

  Close
• Citizen Service Numbers (Netherlands)

  This dictionary detects citizen service numbers (BSN) from the Netherlands.

  The popular format for a citizen service number is a 9-digit number that can be separated by a hyphen after 3 digits.

  An example of a popular format is NNNNNNNNN.

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the citizen service number matches a valid range.	The number formats that can trigger the dictionary are: 132...240774 096195435 25 8662700 2444--52155 041143401
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The citizen service number is in a popular format.	The number formats that can trigger the dictionary are: 096195435 @244452155@ 041143401 The number formats that do not trigger the dictionary are: 132...240774 25 8662700
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The citizen service number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Citizen service number, BSN, Burgersservicenumber, Sofinummer, Persoonsgebonden nummer, Persoonsnummer, or Personal Number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 096195435 @244452155@ 041143401 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 132...240774 25 8662700

  Close
• CNPJ Number (Brazil)*

  This dictionary detects the 14-digit Brazilian National Registry of Legal Entities (CNPJ) number.

  This dictionary uses the Mod 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the CNPJ number matches a valid range.	The number formats that can trigger the dictionary are: 44 455 566 0001 88 44.455/5660001...88 44 -455/ .566 0001...-88 A number format like 44455566000185 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The CNPJ number is in a popular format.	A number format like 44.455.566/0001-88 triggers the dictionary. The number formats that do not trigger the dictionary are: 44455566000188 44 455 566 0001 88 44 -455/ .566 0001...-88 44.455.566/0001 88
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The CNPJ number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, CNPJ, Cadastro Nacional da Pessoa Jurídica, National Registry of Legal Entities, and CNPJ#.	A number format like 44.455.566/0001-88 triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 44455566000188 44 455 566 0001 88 44 -455/ .566 0001...-88 44.455.566/0001 88

  Close
• Corporate Finance Document

  This dictionary detects corporate finance documents, like earnings reports, Form 10-K, etc.

  Zscaler supports only the following document types for corporate finance documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Corporate Legal Document

  This dictionary detects corporate legal documents, like LLC operational agreements, Secretary of State forms, etc.

  Zscaler supports only the following document types for corporate legal documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Corporate Number (Japan)

  This dictionary detects Corporate Numbers from Japan.

  You can modify the Confidence Score Threshold:

  • Low: The dictionary counts an instance as a violation if it matches a valid range.
  • Medium: The dictionary counts an instance as a violation if:
    • The requirements of Low Confidence are met.
    • The Corporate Number is in a popular format.
  • High: The dictionary counts an instance as a violation if:
    • The requirements of Medium Confidence are met.
    • The Corporate Number is accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Court Document

  This dictionary detects court documents, like attorney forms, witness subpoenas, etc.

  Zscaler supports only the following document types for court documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote).The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Credentials and Secrets

  This dictionary allows you to select one or more credentials and secrets dictionaries (i.e., tokens, keys, passwords, etc.) for the following:

  • Amazon MWS Auth Token
  • Git Token
  • GitHub Token
  • Google API Key
  • Google OAuth Access Token
  • Google OAuth ID
  • JWT Token
  • PayPal Braintree Access Token
  • Picatic API Key
  • Private Key
  • SendGrid API Key
  • Slack Access Token
  • Slack Webhook
  • Square Access Token
  • Square OAuth Secret
  • Stripe API Key

  See image.

  

  Close

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Medium	This dictionary counts an instance as a violation if any of the selected sensitive credentials and secrets are matched.
  High	This dictionary counts an instance as a violation if any of the selected sensitive credentials and secrets are matched with a custom high confidence phrase. If you set the Confidence Score Threshold to High, you must specify at least one Custom High Confidence Phrase.

  Close
• Credit Cards

  This dictionary detects content related to credit card numbers.

  When you add the Credit Cards dictionary to a DLP engine, you must configure the match count. To learn more, see Configuring the Match Count.

  See image.

  

  Close

  The popular format for a credit and debit card number is a range from 12-19 numeric digits, separated after every 4 digits by a delimiter that is checked for similarity and is correctly spaced. The delimiter can be a slash, period, hyphen, or space.

  The following are examples of popular formats:

  • AMEX (15 digits)
    • NNNN<delimiter>NNNNNN<delimiter>NNNNN
  • Diner's Club (14 digits)
    • NNNN<delimiter>NNNNNN<delimiter>NNNN
  • China UnionPay (16 digits beginning with 62 or 60)
    • 62NNNNNNNNNNNNN
    • 62NNNNNNNNNNNNNNNNN
  • Maestro Debit Card (16–19 digits starting with 50, 56, 57, 58, 6013, 62, 63, or 67)
    • 50NN-NNNN-NNNN
    • 6013 NNNN NNNN
  • Other Credit Cards (16 digits)
    • NNNN<delimiter>NNNN<delimiter>NNNN<delimiter>NNNN
    • NNNN NNNN NNNN NNNN

  This dictionary uses the Luhn checksum.

  You can also specify whether to include or exclude specific Bank Identification Numbers (BIN) in credit card dictionaries. You can configure up to 512 BIN values. To learn more, see Editing Predefined DLP Dictionaries.

  To use this dictionary to detect credit card numbers, the following guidelines apply:

  These guidelines might not apply to certain use cases. You can adjust your dictionary configurations according to your DLP needs.

  • To catch a small exfiltration of numbers in documents:
    • Set a Confidence Score Threshold of High for the dictionary.
    • When adding a dictionary to an engine, set a low value for the dictionary's match count.
  • To catch a large exfiltration of numbers in spreadsheets or other file types:
    • Set a Confidence Score Threshold of Medium for the dictionary.
    • When adding a dictionary to an engine, set a high value for the dictionary’s match count.
  • In general, configuring a dictionary with a Low Confidence Score Threshold and a low match count value results in too many false positives. Zscaler recommends setting a Confidence Score Threshold of High when its match count value is low.
  • Rich Text Format (RTF) files contain formatting code that can mimic credit card and social security numbers, affecting when a DLP rule is triggered. Plain text files do not contain this formatting code, therefore the DLP rule triggers as expected. So that the DLP policy triggers if confidential numbers are leaked in RTF files, do one of the following steps:
    • Set any value greater than 1 for the dictionary’s match count in the engine.
    • Set a Confidence Score Threshold value of High.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if: The credit card number matches the length of at least one provider. The credit card number can be validated by Luhn checksum. The credit card number can contain a period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed.	The number formats that can trigger the dictionary are: 491-6710478214413 491-67...104782 14413 47 1697--47625799...21 36839-32 9773518 53721 653983--86508 4716 7361 13842732 (Valid length and can be validated by Luhn checksum) The number formats that do not trigger the dictionary are: aa5269946762375011aa 4716 7361 1384 2731 (Cannot be validated by the Luhn checksum)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The credit card number is in a popular format. The credit card number, the length, and starting range of the number match that of the credit card providers. The credit card number must use the same delimiters for the full number. Plus, the delimiters must be equally spaced.	The number formats that can trigger the dictionary are: 4716-9747-6257-9921 5372165398386508 4716 7361 1384 2732 The number formats that do not trigger the dictionary are: 4916:7104-7821 4413 49167-104-7821-4413 a3683:9329:773-518a
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The credit card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Amex, MasterCard, Visa, CVV Code, CCV Number, select card type, Discover, Diners Club, jcb, pay with checking account, pay check money order, credit card number, card holder name, or expiration date.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 4716-9747-6257-9921 5269:9467:6237:5011 5372165398386508 4716 7361 1384 2732 A number like a3683:9329:773-518a does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  If a violation is enough to trigger a DLP policy and you have configured a DLP email notification, your auditors receive an email. If you have also included the ${DLPTRIGGERS} macro, the email includes what content triggered the violation.

  See image.

  

  Close

  Close
• Diseases Information

  This dictionary uses phrase matching to detect content related to diseases information. It does not use a checksum.

  You can specify an Action to configure how the dictionary evaluates matching disease names:

  • Count All: The dictionary counts all matches of the disease name, including identical disease names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the disease name toward the match count only once, regardless of how many times the disease name appears.

  Close
• Driver’s License (United States)

  This dictionary allows you to select driver's license dictionaries for one or more of the 50 U.S. states plus the District of Columbia.

  See image.

  

  Close

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Medium	This dictionary counts an instance as a violation if the selected driver's license is matched.
  High	The requirements of Medium Confidence are met and the driver's license is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Drivers License, Drivers License number, DL#, or 2-letter state codes (e.g., CA for California or WA for Washington).

  The Proximity Length field appears when the dictionary’s Confidence Score Threshold is High. The proximity length defines how close a high confidence phrase must be to an instance of the pattern (that the dictionary detects) to count as a match. The phrase can be located in any direction from the pattern within the document. Enter a value from 0–10,000 bytes. A proximity length of 0 disables this option (i.e., the phrase can be any distance from the pattern).

  Close
• Drugs Information

  This dictionary uses phrase matching to detect content related to drug information. It does not use a checksum.

  You can specify an Action to configure how the dictionary evaluates matching drug names:

  • Count All: The dictionary counts all matches of the drug name, including identical drug names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the drug name toward the match count only once, regardless of how many times the drug name appears.

  Close
• Financial Statements

  This dictionary detects content related to financial statements. It detects a financial document based on machine learning which clusters keywords that typically occur in financial documents. Examples of keywords are Accounts Payable, Common Stock, Current Liabilities, and so on.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• First Names (Japan)

  This dictionary detects content related to First Names from Japan.

  This dictionary does not use a checksum.

  The following are examples of acceptable formats for Japanese First Names:

  • 三郎
  • 清子
  • 美代子
  • 四郎

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Fiscal Code (Italy)

  This dictionary detects Italian fiscal codes from Italy.

  The popular format for an Italian fiscal code is a 16-character alphanumeric code that has no delimiters.

  An example of a popular format is FFF-NNN-YYMDD-RRRRC. It has the following structure:

  • FFF: Represents the first three letters of the last name.
  • NNN: Represents the first three letters of the first name.
  • YYMDD: Represents the date of birth.
    • YY: Represents the year.
    • M: Represents the month. It is represented by a letter that maps to a month.
    • DD: Represents the date.
  • RRRR: Represents the town of birth. It is represented by alphnumeric characters.
  • C: Represents the checksum value. It is represented by a letter.

  This dictionary uses the Mod 26 Check Digit checksum (maps to a letter A-Z).

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The Fiscal Code (Italy) dictionary is a regex based dictionary. It is more strict in formatting, even in low confidence. The special characters must be in the right positions. There could be a mix of different special characters, and there may even be consecutive special characters at the right positions, but they cannot be in the wrong positions. The dictionary counts an instance as a violation if the Italian fiscal code matches a valid range. The Italian fiscal code can contain only a period, hyphen, or space as delimiters. Multiple periods are allowed for a low confidence score.	The number formats that can trigger the dictionary are: RSS MRA 70A41 F205Z KAYSAN92D03L246A RSS MRA- ... 70A41F205Z KAYSAN 92D03L246 ... A The number formats that do not trigger the dictionary are: RSS MRA70A41F20 5Z KAYSAN9 2D03L246A
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Italian fiscal code is in a popular format.	The number formats that can trigger the dictionary are: RSS MRA 70A41 F205Z KAYSAN92D03L246A The number formats that do not trigger the dictionary are: RSS MRA- ... 70A41F205Z KAYSAN 92D03L246 ... A
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Italian fiscal code is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, codice fiscal, repubblica italiana, Italian fiscal code, or Italian tax code.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: RSS MRA 70A41 F205Z KAYSAN92D03L246A The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: RSS MRA- ... 70A41F205Z KAYSAN 92D03L246 ... A

  Close
• Full Names (Japan)

  This dictionary detects content related to Full Names from Japan.

  This dictionary does not use a checksum.

  For example, if the First Name is 富市, and the Last Name is 村山, the following are examples of acceptable formats for Japanese Full Names:

  • 村山富市 (no delimiters)
  • 村山 富市 (spaces as delimiter)
  • 村山　富市 (unicode space as delimiter)
  • 村山,富市 (ASCII comma as delimiter)

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Gambling

  This dictionary detects content related to gambling. It detects documents that have keywords related to both online and physical gambling and betting.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Identity Card Number (China)

  This dictionary detects Resident Identity Card numbers from China.

  The popular format for a Resident Identity Card number is an 18-character number, but the last character can either be a digit or the character X (case insensitive).

  The following are examples of popular formats:

  • NNNNNNNNNNNNNNNNNX
  • NNNNNNNNNNNNNNNNNN

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Resident Identity Card number matches a valid range.	The number formats that can trigger the dictionary are: 36--2331198--00322524X 3607 2719830115027X 3425011992...07064058 A number format like 1404011998vv05055835 does not trigger the dictionary. It contains alphabetical characters
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Resident Identity Card number is in a popular format.	The number formats that can trigger the dictionary are: 36072719830115027X 342501199207064058 The number formats that do not trigger the dictionary are: a36233119800322524X (Starts with a character) a1404011998vv05055835 (Starts with a character)
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Resident Identity Card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Chinese identity card number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 36233119800322524X 36072719830115027X 342501199207064058 140401199805055835

  Close
• Identity Card Number (Hong Kong)

  This dictionary detects Hong Kong identity card (HKID) numbers from Hong Kong.

  The popular format for an HKID number is 8 or 9 alphanumeric characters without delimiters. It starts with either one or two alphabet letters, followed by 6 random digits, and ends with a checksum character that must be enclosed in parentheses. The checksum character can be either a digit or the letter "A" (case insensitive). For example, P553722(7), MR427885(6), and FK057839(A).

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the HKID number matches a valid range. HKID is a regex-based dictionary, and is strict in formatting, even in low confidence. No special characters are allowed other than parentheses. If the parentheses are not balanced, it does not trigger the dictionary. There can be only a single pair of parentheses. Nested parentheses do not trigger the dictionary.	The number formats that can trigger the dictionary are: P5537227 P553722(7) P553722(A) The number formats that do not trigger the dictionary are: P553722-7- P553722(7 P5537227) P553722((7))
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The HKID number is in a popular format.	The number format that can trigger the dictionary is P553722(7). The number format that does not trigger the dictionary is P5537227.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The HKID number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Hong Kong Identity Card, HKIC, HKID, Identity Card, or Hong Kong Permanent Resident ID Card.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: P553722(7) P553722(A)

  Close
• Identity Card Number (Malaysia)

  This dictionary detects National Identity Card (MyKad) numbers from Malaysia.

  The popular format for a MyKad number is a 12-digit number. The first group of numbers is the date of birth (YYMMDD). The second group of numbers (SS) represents the place of birth of the holder. The states (01-13), the federal territories (14-16), or the country of origin (60-85). The last group of numbers (###G) is a serial number in an unidentified pattern which is randomly generated. The last digit (G) is an odd number for a male and an even number for a female.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the MyKad number matches a valid range.	The number formats that can trigger the dictionary are: 13...0125281813 261231129312 170 726145727 @240921167814@ 16022--5148988 A number format like 1007192dd33724 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The MyKad number is in a popular format.	The number formats that can trigger the dictionary are: 261231129312 @240921167814@ The number formats that do not trigger the dictionary are: 13...0125281813 170 726145727 16022--5148988 A100719233724A
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The MyKad number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, mykad, Malaysian nric, or mypr.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 261231129312 @240921167814@ 100719233724 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 13...0125281813 170 726145727 16022--5148988

  Close
• Identity Card Number (Thailand)

  This dictionary detects National Identity Card numbers from Thailand.

  The popular format for a National Identity Card number is a 13-digit string in the format of N-NNNN-NNNNN-NN-N. In the Popular Format Checking process, it checks for 4 delimiters.

  This dictionary uses the Luhn variation checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the National Identity Card number matches a valid range. The National Identity Card number can contain only a period, hyphen, or space as delimiters.	The number formats that can trigger the dictionary are: @9082208241537@ 1-1964-43062-03-2 5-40 42-291 43-14-9 8161427577191 3-5887-693...20-66-7 A number format like 070ww7389561584 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The National Identity Card number is in a popular format.	The number formats that can trigger the dictionary are: @9082208241537@ 1-1964-43062-03-2 8161427577191 The number formats that do not trigger the dictionary are: 070ww7389561584 5-40 42-291 43-14-9 3-5887-693...20-66-7
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The National Identity Card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, thailand identity card number, thailand national, date issue, or date expiry.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: @9082208241537@ 1-1964-43062-03-2 8161427577191 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 07ss07389561584 5-4042-291cc43-14-9

  Close
• Illegal Drugs

  This dictionary detects content related to illegal drugs. It detects the names of illegal drugs such as Cocaine, Heroin, Ketamine, and so on.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Immigration Document

  This dictionary detects immigration documents, like passport renewal forms, I-485, I-856, I-907, etc.

  Zscaler supports only the following document types for immigration documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Individual Taxpayer Registry ID (Brazil)

  This dictionary detects Individual Taxpayer Registry ID numbers (CPF) from Brazil. If you use an Exact Data Match (EDM) Index Template to try and detect a CPF number, the format ddd.ddd.ddd-dd does not work because '.' (period) is an EDM delimiter. However, if you include this dictionary in a rule, the format ddd.ddd.ddd-dd is detected.

  The popular format for an Individual Taxpayer Registry ID number is different for individuals versus legal persons. For individuals, it is an 11-digit number with the last two numbers being the result of an arithmetic operation from the 9 previous ones. For legal persons, it is a 14-digit string formatted as XX.XXX.XXX/XXXX-XX. The first 8 digits identify the company, the four digits after the slash identify the branch or subsidiary, and the last 2 digits are the result of an arithmetic operation from the previous ones.

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Individual Taxpayer Registry ID number matches a valid range. The Individual Taxpayer Registry ID number can contain: A period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. An alphabetical boundary.	The number formats that can trigger the dictionary are: 088.258.987-38 103.015.819-32 The number formats that do not trigger the dictionary are: 989.786.5232-27 286.648.5a71-80 345.543.66@5-02
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Individual Taxpayer Registry ID number is in a popular format. The Individual Taxpayer Registry number can contain a non-alphanumeric boundary. It cannot be bound by alphabetical characters.	The number formats that can trigger the dictionary are: 286.648.571-80 088.258.987-38 @345.543.665-02@ 103.015.819-32 A number format like 989.786.523-2--7 does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Individual Taxpayer Registry ID number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Natural Persons Register or Registration Number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 286.648.571-80 088.258.987-38 @345.543.665-02@ 103.015.819-32 A number format like 989.786.523-2--7 does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Insurance Document

  This dictionary detects insurance documents, like employee insurance, home insurance, commercial insurance, medical insurance, etc.

  Zscaler supports only the following document types for insurance documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• International Bank Account Number (IBAN)*

  This dictionary allows you to select IBAN dictionaries for one or more of the following countries:

  • Andorra (AD)
  • Austria (AT)
  • Bosnia (BA)
  • Belgium (BE)
  • Bulgaria (BG)
  • Switzerland (CH)
  • Cyprus (CY)
  • Czechia (CZ)
  • Germany (DE)
  • Denmark (DK)
  • Estonia (EE)
  • Spain (ES)
  • Finland (FI)
  • Faroe Islands (FO)
  • France (FR)
  • United Kingdom (GB)
  • Gibraltar (GI)
  • Greenland (GL)
  • Greece (GR)
  • Croatia (HR)
  • Hungary (HU)
  • Ireland (IE)
  • Israel (IL)
  • Iceland (IS)
  • Italy (IT)
  • Liechtenstein (LI)
  • Lithuania (LT)
  • Luxembourg (LU)
  • Latvia (LV)
  • Monaco (MC)
  • Montenegro (ME)
  • Malta (MT)
  • Netherlands (NL)
  • North Macedonia (MK)
  • Norway (NO)
  • Poland (PL)
  • Portugal (PT)
  • Romania (RO)
  • Serbia (RS)
  • Sweden (SE)
  • Slovenia (SI)
  • Slovakia (SK)
  • San Marino (SM)
  • Tunisia (TN)
  • Turkey (TR)

  See image.

  

  Close

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Medium	This dictionary counts an instance as a violation if any of the selected IBAN numbers are matched.
  High	This dictionary counts an instance as a violation if any of the selected IBAN numbers are matched by any of the dictionary’s default or custom high confidence phrases.

  Close
• Invoice Document

  This dictionary detects invoice documents, like Bill of Sale forms, purchase orders, etc.

  Zscaler supports only the following document types for invoice documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Last Names (Japan)

  This dictionary detects content related to Last Names from Japan.

  This dictionary does not use a checksum.

  The following are examples of acceptable formats for Japanese Last Names:

  • 村山
  • 佐久間
  • 平山
  • 五十嵐
  • 佐藤
  • 鈴木

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Legal Document

  This dictionary detects legal documents, like living wills, name change certificates, etc.

  Zscaler supports only the following document types for legal documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Medical Document

  This dictionary detects medical documents, like medical consent forms, HIPAA forms, medical record forms, etc.

  Zscaler supports only the following document types for medical documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Medical Information

  This dictionary detects content related to medical information. It detects the drug and disease names based on the ICD 10 code.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Medicare Numbers (Australia)

  This dictionary detects Medicare numbers from Australia.

  The popular format for a Medicare number is an 11-digit number. In the Popular Format Checking process, the Mod 10 checksum is performed and the number has to pass the checksum.

  This dictionary uses the Mod 10 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Medicare number matches a valid range.	The number formats that can trigger the dictionary are: @6014812406@ 59...89916...497 39--388984 01 A number format like 24//2877813//2 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Medicare number is in a popular format.	The number formats that can trigger the dictionary are: @6014812406@ 5989916497 3938898401 A number format like A2428778132B does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Medicare number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, bank account details, medicare payments, mortgage account, bank payments, information branch, credit card loan, department human services, medicare, or medi care.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: @6014812406@ 5989916497 3938898401 A number format like A2428778132B does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Mexico Unique Population Registration Code*

  This dictionary detects Mexico Unique Population Registration Code numbers.

  This dictionary uses the Mod 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Mexico Unique Population Registration Code number matches a valid range.	The number formats that can trigger the dictionary are: HEGG.560427MVZRRL04 HEGG 560427MVZRRL04 HEGG-560427MVZRRL04 HEGG.- 560427MVZRRL04 HEGG560427 MVZRRL04 HEGG560427.MVZRRL04 HEGG560427-MVZRRL04 HEGG560427 -MVZRRL04 HEGG 560427-MVZRRL04 The number formats that do not trigger the dictionary are: HEGG560427MVZRRL03 (bad checksum) HE GG560427MVZRRL04 (unpopular format; doesn't match regex) HEGG560427MVZRRL 04 (unpopular format; doesn't match regex) HE GG 560427 MVZRRL 04 (unpopular format; doesn't match regex)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Mexico Unique Population Registration Code number is in a popular format.	The number formats that can trigger the dictionary are: HEGG560427MVZRRL04 hEGG560427MVZRRL04 hegg560427mvzrrl04 A number format like HEGG-560427MVZRRL04 does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Mexico Unique Population Registration Code number is accompanied by any of the dictionary’s default high confidence phrases. For example, Clave Única de Registro de Población, CURP, clave única, ClaveÚnica#, and clavepersonalIdentidad#.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: HEGG560427MVZRRL04 hEGG560427MVZRRL04 hegg560427mvzrrl04 A number format like HEGG-560427MVZRRL04 does not trigger the dictionary.

  Close
• My Number (Japan)

  This dictionary detects My Numbers (also referred to as Individual Numbers) from Japan.

  You can modify the Confidence Score Threshold:

  • Low: The dictionary counts an instance as a violation if it matches a valid range.
  • Medium: The dictionary counts an instance as a violation if:
    • The requirements of Low Confidence are met.
    • The My Number is in a popular format.
  • High: The dictionary counts an instance as a violation if:
    • The requirements of Medium Confidence are met.
    • The My Number is accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Names (Canada)

  This dictionary detects content related to names from Canada.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	The dictionary counts an instance as a violation if the content contains either a first name or a last name that is in the dictionary.
  Medium	The dictionary counts an instance as a violation if the content contains a first name and last name that is in the dictionary.

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Names (Spain)

  This dictionary detects content related to names from Spain.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	The dictionary counts an instance as a violation if the content contains either a first name or a last name that is in the dictionary.
  Medium	The dictionary counts an instance as a violation if the content contains a first name and last name that is in the dictionary.

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• Names (US)

  This dictionary detects content related to names from the United States. It detects first and last names.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	The dictionary counts an instance as a violation if the content contains either a first name or a last name that is in the dictionary.
  Medium	The dictionary counts an instance as a violation if the content contains a first name and last name that is in the dictionary.

  You can also specify an Action to configure how the dictionary evaluates matching names:

  • Count All: The dictionary counts all matches of the name, including identical names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the name toward the match count only once, regardless of how many times the name appears.

  Close
• National Economic Registry Number (Poland)*

  This dictionary detects the 9- or 14-digit Polish National Economic Registry Number (REGON) number.

  This dictionary uses the Mod 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the REGON number matches a valid range.	The number formats that can trigger the dictionary are: 13 3456783 13345678 3 13 345678-3 13 345678312340 13 34567831234-0 13 4567831234 /-0 A number format like 133456788 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The REGON number is in a popular format.	The number formats that can trigger the dictionary are: 133456783 13 345678 3 13345678312340 13/34567831234/0 The number formats that do not trigger the dictionary are: 13 3456783 13345678 3 13 345678-3 13 345678312340 13 34567831234-0 13 4567831234 /-0
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The REGON number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, regon.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 133456783 13 345678 3 13345678312340 13/34567831234/0 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 13 3456783 13345678 3 13 345678-3 13 345678312340 13 34567831234-0 13 4567831234 /-0

  Close
• National Health Index Number (New Zealand)

  This dictionary detects New Zealand National Health Index Numbers (NZNHIN).

  NZNHINs have two formats, both of which are undelimited with no special characters in between:

  • Old format: AAANNNC
  • New format: AAANNAC

  The old format uses Modulo 11 checksum; the new format uses the Modulo 24 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NZNHIN number matches a valid range.	A number format like WLD-9413 can trigger the dictionary. The number formats that do not trigger the dictionary are: WL-D9413 (doesn't match regex) WLD9-413 (doesn't match regex) WLD94-13 (doesn't match regex)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NZNHIN number is in a popular format.	The number formats that can trigger the dictionary are: WLD9413 aDh48zj A number format like WLD-9413 does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The NZNHIN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, National Health Index Number, National Health Index Num, NHI number, or NHI#.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: WLD9413 aDh48zj A number format like WLD-9413 does not trigger the dictionary.

  Close
• National Health Service Number (UK)

  This dictionary detects National Health Service (NHS) numbers from the United Kingdom.

  The popular format for an NHS number is a 10-digit number formatted as NNN <delimiter>NNN<delimiter>NNNN. The delimiters can be a period, hyphen, or space.

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NHS number matches a valid range. The NHS number can contain: A period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. An non-alphanumeric boundary.	The number formats that can trigger the dictionary are: 566 8018326 5431043544 808619--0226 5878649888 130323...3851
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NHS number is in a popular format. The NHS number can contain a non-alphanumeric boundary. It cannot be bound by alphabetical characters. The NHS number must use the same delimiters for the full number.	The number formats that can trigger the dictionary are: @5878649888@ 1303233851 The number formats that do not trigger the dictionary are: 5@668018326 543!1043!544 A8086190226A
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The NHS number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, NHS Number or National Health Services Number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 5668018326 5431043544 A8086190226A @5878649888@ 1303233851

  Close
• National Identification Card Number (Taiwan)

  This dictionary detects National Identification Card numbers from Taiwan.

  The popular format for a National Identification Card number is a 10-digit string that contains one letter and 9 digits. It can also contain a period, hyphen, or space as delimiters.

  This dictionary uses the Luhn checksum but the leading alphabet letter gets converted to a numerical equivalent.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the National Identification Card number matches a valid range. The National Identification Card number can contain a period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. The National Identification Card number can not contain a delimiter between the first alphabet letter and the next number.	The number formats that can trigger the dictionary are: D146. 665-645 --I145639659 !!F172388317VV Z258578175DD The number formats that do not trigger the dictionary are: Z...222063149 !!X14234881733
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The National Identification Card number is in a popular format. The National Identification Card number can only have a non-alphanumeric character for the end delimiter.	The number formats that can trigger the dictionary are: I145639659 !!X142348817@@ Z258578175 The number formats that do not trigger the dictionary are: Z...222063149 F172388317VV
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The National Identification Card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, taiwanese national identification card number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: I145639659 Z222063149 !!X142348817@@ Z258578175 A number format like F172388317VV does not trigger the dictionary if accompanied by any of the dictionary's default or custom high confidence phrases.

  Close
• National Identification Number (France)

  This dictionary detects National Institute of Statistics and Economic Studies (INSEE) numbers from France.

  The popular format for an INSEE number is a 15-digit number with the first 13 digits and last 2 digits separated by a space.

  This dictionary uses the Mod by 97 checksum. The last two digits are check digits.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the INSEE number matches a valid range. The INSEE number can contain: A period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. An alphabetical boundary.	The number formats that can trigger the dictionary are: @113103115324014@ 127...1139450--100 58 #23--0036713141980#
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The INSEE number is in a popular format. The INSEE number can contain: A non-alphanumeric boundary. Only one space between N(13) and N(2). The INSEE number can not contain a period or a hyphen as a delimiter.	The number formats that can trigger the dictionary are: 2430966952225 59 1160568707959 73 %1271139450100 58$ #2300367131419 80# The number formats that do not trigger the dictionary are: A138057773213587c 113103115324014
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The INSEE number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, insee, national id, national identification, social security number, social security code, and social insurance number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 2430966952225 59 1160568707959 73

  Close
• National Identification Number (Poland)

  This dictionary detects National Identification (PESEL) numbers from Poland.

  The popular format for a PESEL number is an 11-digit number. The PESEL number has the form of YYMMDDZZZXQ, where YYMMDD is the date of birth (with century encoded in month field), ZZZX is the personal identification number where X codes the sex (even number for females and odd number for males), and Q is a check digit which is used to verify whether a given PESEL is correct or not.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the PESEL number matches a valid range.	The number formats that can trigger the dictionary are: .14.1017...04491 -82-12270--5195- 971 0050 7364 The number formats that do not trigger the dictionary are: 2302130d3028 (55111304237)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The PESEL number is in a popular format.	The number formats that can trigger the dictionary are: 97100507364 @23021303028$ The number formats that do not trigger the dictionary are: .14.1017...04491 -82-12270--5195- A55111304237G
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PESEL number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, pesel liczba or peselliczba.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 14101704491 82122705195 97100507364 @23021303028$ 55111304237

  Close
• National Identification Number (Spain)

  This dictionary detects National Identity Card numbers (DNI) from Spain.

  The popular format for a National Identity Card number is a 9-character number with 8 digits and one letter for security.

  This dictionary uses the Mod by 23 checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the National Identity Card number matches a valid range. The National Identity Card number can contain: A period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed. An alphanumeric character as a lead boundary.	The number formats that can trigger the dictionary are: 20---222624N a22369319W The number formats that do not trigger the dictionary are: 0 1311947@G 33052840-T
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The National Identity Card number is in a popular format. The National Identity Card number can contain an alphabetical character as an ending boundary. The National Identity Card number can not have a delimiter between the last alphabet character and the number.	The number formats that can trigger the dictionary are: 01311947G 20222624N 33052840T A number format like a22369319W does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The National Identity Card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, national identification number, national identity number, insurance number, personal identification number, national identity, personal identity no, or unique identity number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 20222624N 22369319W 33052840T A number format like d01311947G does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• National Insurance Number (UK)

  This dictionary detects National Insurance Numbers (NINO) from the United Kingdom.

  The popular format for a NINO is a 9-character number. The format of the number is two prefix letters, 6 digits, and one suffix letter. Either of the first two prefix letters cannot be D, F, I, Q, U, or V. The second letter cannot also be O. The prefixes BG, GB, NK, KN, TN, NT, and ZZ are not allocated.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NINO matches a valid range.	The number formats that can trigger the dictionary are: AA-123456.C @#AA876589C@ BB...123456...D A number format like RR65------3456 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NINO is in a popular format.	The number formats that can trigger the dictionary are: AA123456C AA876589C The number formats that do not trigger the dictionary are: BB123456...D AA-127456.G
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The NINO is accompanied by any of the dictionary’s default high confidence phrases. For example, national insurance number or national insurance.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: AA123465C RK765432C The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: rh123456U AA56432D

  Close
• National Provider Identifier

  This dictionary detects National Provider Identifier (NPI) numbers from the United States.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NPI number matches a valid range.	The number formats that can trigger the dictionary are: 123 456 7893 12.345-6789 3 1 -23/ .456 78...- 9 3 A number format like 1234567894 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NPI number is in a popular format.	A number format like 1234567893 can trigger the dictionary. The number formats that do not trigger the dictionary are: 123 456 7893 12.345-6789 3 1 -23/ .456 78...- 9 3
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The NPI number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, national provider identifier number, national provider identifier, and npi.	A number format like 1234567893 can trigger the dictionary. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 123 456 7893 12.345-6789 3 1 -23/ .456 78...- 9 3

  Close
• NDC Number (Package)

  This dictionary detects content related to National Drug Code (NDC) package codes.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if it matches a key in the hash table.	The number formats that can trigger the dictionary are: 0002080001 0990-925739 10014001-08 99528---606--45 The number formats that do not trigger the dictionary are: 000-20800-01 (first hyphen at wrong position) 0990-925-739 (second hyphen at wrong position) 0002-08-00-01 (too many non-adjacent hyphens) 99528---6-06--45 (too many non-adjacent hyphens)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NDC package code is in a popular format.	The number formats that can trigger the dictionary are: 0002-0800-01 10014-001-08 86157-0014-1 The number formats that do not trigger the dictionary are: 0002080001 0990-925739 10014001-08 99528---606--45

  You can also specify an Action to configure how the dictionary evaluates matching package codes:

  • Count All: The dictionary counts all matches of the package code, including identical codes, toward the match count.
  • Count Unique: The dictionary counts each unique match of the package code toward the match count only once, regardless of how many times the code appears.

  Close
• NDC Number (Product)

  This dictionary detects content related to National Drug Code (NDC) product codes.

  This dictionary does not use a checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if it matches a key in the hash table.	The number formats that can trigger the dictionary are: 00020800 0990--9257 The number formats that do not trigger the dictionary are: 000-20800 (hyphen at wrong position) 00020-800 (hyphen at wrong position) 0990-925-7 (too many non-adjacent hyphens)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NDC package code is in a popular format.	The number formats that can trigger the dictionary are: 0002-0800 10014-001 86157-0014 The number formats that do not trigger the dictionary are: 00020800 0990--9257

  You can also specify an Action to configure how the dictionary evaluates matching product code:

  • Count All: The dictionary counts all matches of the product code, including identical codes, toward the match count.
  • Count Unique: The dictionary counts each unique match of the product code toward the match count only once, regardless of how many times the code appears.

  Close
• NRIC Numbers (Singapore)

  This dictionary detects National Registration Identity Card Numbers (UIN and FIN) from Singapore.

  The popular format for a National Registration Identity Card number is a 9-character number that includes 7 digits, with the first character being either S, T, F, or G, and the last character is a checksum calculated with respect to the first letter and the 7 digits.

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the National Registration Identity Card number matches a valid range. The National Identity Card number can contain a period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed.	The number formats that can trigger the dictionary are: T156--2546A G545...8130R aS8879619Ed The number formats that do not trigger the dictionary are: F2d397111U T0v75 8616C
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The National Registration Identity Card number is in a popular format. The National Registration Identity Card number can not contain any delimiters. The boundary check is not performed for the National Registration Identity Card number as both the start and end have a character.	The number formats that can trigger the dictionary are: T1562546A G5458130R aS8879619Ed The number formats that do not trigger the dictionary are: F2d397111U T0v758616C
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The National Registration Identity Card number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, nric, national registration identity card, guin, fin, passport number, or birth certificate.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: T1562546A G5458130R F2397111U T0758616C aS8879619Ed

  Close
• Passport Number (Asia)*

  This dictionary allows you to select passport number dictionaries for one or more of the following countries.

  • China (CN)
  • Japan (JP)
  • South Korea (KR)
  • Malaysia (MY)
  • Philippines (PH)
  • Singapore (SG)
  • Taiwan (TW)
  • Turkey (TR)

  See image.

  

  Close

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Medium	This dictionary counts an instance as a violation if any of the selected passport numbers are matched.
  High	This dictionary counts an instance as a violation if any of the selected passport numbers are matched by any of the dictionary’s default or custom high confidence phrases.

  Close
• Passport Number (European Union)

  This dictionary allows you to select passport number dictionaries for one or more European Union countries.

  See image.

  

  Close

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Medium	This dictionary counts an instance as a violation if the selected country's passport number is matched.
  High	The requirements of Medium Confidence are met and the passport number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Passport, Passport number, passport#.

  Close
• Personal Identification Number (Croatia)*

  This dictionary detects Personal Identification Number (Croatia) numbers.

  This dictionary uses the ISO/IEC 7064, Mod 11,10 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the PIN (Croatia) number matches a valid range.	The number formats that can trigger the dictionary are: hr 94577403194 hr.- 12345678911 HR..9942692209 6 The number formats that do not trigger the dictionary are: HR69435151531 (bad checksum) H R69435151538 (unpopular format; doesn't match regex) HR69435 151538 (unpopular format; doesn't match regex) 8684 9310 961 (unpopular format; doesn't match regex)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The PIN (Croatia) number is in a popular format.	The number formats that can trigger the dictionary are: HR69435151531 HR 69435151531 24631579813 The number formats that do not trigger the dictionary are: 94 577 403194 hr.-12345678911 HR69435 151538 hr.- 12345678911
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PIN (Croatia) number is accompanied by any of the dictionary’s default high confidence phrases. For example, OIB, Osobni identifikacijski broj, and Personal identification number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: HR69435151531 HR 69435151531 24631579813 The number formats that do not trigger the dictionary are: 94 577 403194 hr.-12345678911 HR69435 151538 hr.- 12345678911

  Close
• Real Estate Document

  This dictionary detects real estate documents, like personal or commercial lease agreements, property buying or selling agreements, etc.

  Zscaler supports only the following document types for real estate documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Resident Registration Number (Korea)

  This dictionary detects Resident Registration Numbers (RRN) from South Korea.

  The popular format for a RRN is a 13-digit number with each digit providing specific information.

  This dictionary uses the Luhn variation checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the RNN matches a valid range. The RNN can contain: A period, hyphen, or space as delimiters. Multiple periods are allowed. An alphabetical boundary.	The number formats that can trigger the dictionary are: 9 70403-2966211 780220-1296377 #840...719-2145299@ D921009-5664079D A number format like 720590-2208919 does not trigger the dictionary because it fails the popular format check.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The RRN is in a popular format. The RRN can contain: A non-alphanumeric boundary. Only a hyphen as a delimiter.	The number formats that can trigger the dictionary are: 970403-2966211 !780220-1296377@ The number formats that do not trigger the dictionary are: 720590-2208919 #840...719-2145299@ D921009-5664079D
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The RNN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, korean resident registration number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 970403-2966211 !780220-1296377@ 890320-1104929 840719-2145299 921009-5664079 940219-5027845

  Close
• Resume Document

  This dictionary detects resume documents.

  Zscaler supports only the following document types for resume documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Salesforce.com Data

  This dictionary detects content related to Salesforce.com data. It detects keywords that are embedded in a Saleforce report, such as Copyright and Saleforce.com.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Self-Harm & Cyberbullying

  The Self-Harm and Cyberbulling dictionary is intended for use by K-12 educators. If you’re an educator, contact your Zscaler Account team to enable this dictionary.

  This dictionary detects when students use public search engines to search for self-harm-related content, or when they post bullying-related content on social media.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Social Insurance Numbers (Canada)

  This dictionary detects Social Insurance Numbers (SIN) from Canada.

  The popular format for an SIN is a 9-digit number, separated at the 3rd and 6th digits by a delimiter. The delimiter can be a period, hyphen, or space.

  The following are examples of popular formats:

  • NNNNNNNNN
  • NNN-NNN-NNN

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the SIN matches a valid range. The SIN can contain only a period, hyphen, or space as delimiters. Multiple periods, hyphens, and spaces are allowed.	The number formats that can trigger the dictionary are: CC036964 310CC @054120 373$ @620301 192% AA021574 728 A number format like 650------72...aa...7266 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The SIN is in a popular format. The SIN can contain delimiters only after the 3rd and 6th digits. The delimiters must be the same character. It cannot contain different delimiters. The SIN can have a boundary but the characters immediately before and after the SIN cannot be alpha-numeric.	The number formats that can trigger the dictionary are: 054120373 650-727-266 The number formats that do not trigger the dictionary are: @620 301 192 AA021574 728 650-72.7266
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The SIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, social insurance number or national identification number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 036 964 310 054120373 @620 301 192 650-72.7266 369 893 144 A number format like AA021574 728 does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Social Security Number (Austria)

  This dictionary detects Social Security numbers (ATSSN) from Austria.

  This dictionary uses the Modulo 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the ATSSN number matches a valid range.	The number formats that can trigger the dictionary are: 123 701 0180 1237--010180 1 -23/.701 -. 0 ... -180 A number format like 1238010180 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The ATSSN number is in a popular format.	The number formats that can trigger the dictionary are: 1237010180 1237 010180 The number formats that do not trigger the dictionary are: 123 701 0180 1237--010180 1 -23/.701 -. 0 ... -180
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The ATSSN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, sozialversicherungsnummer, Austria SSN, soziale sicherheit kein, sozialversicherungsnummer#, and sozialesicherheitkein#.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 1237010180 1237 010180 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 123 701 0180 1237--010180 1 -23/.701 -. 0 ... -180

  Close
• Social Security Number (Spain)

  This dictionary detects social security numbers from Spain.

  You can modify the confidence score threshold:

  • Low: The dictionary counts an instance as a violation if it matches a valid range.
  • Medium: The dictionary counts an instance as a violation if:
    • The requirements of Low Confidence are met.
    • The social security number is in a popular format.
  • High: The dictionary counts an instance as a violation if:
    • The requirements of Medium Confidence are met.
    • The social security number is accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Social Security Number (Switzerland)

  This dictionary detects social security numbers (AHV) from Switzerland.

  The popular format for an AHV number is a 13-digit number. The AHV number has the form of 756.XXXX.XXXX.XY, where 756 is the ISO 3166-1 code for Switzerland, XXXX.XXXX.X is a random number, and Y is an EAN-13 check digit.

  This dictionary uses the Luhn variation checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the AHV number matches a valid range.	The number formats that can trigger the dictionary are: 756.52--30.6913.27 756.88 85.8748.24 756.3594.2833.18 @7.56.7811.8967.63@ A number format like 756.2...487.40cc93.09 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The AHV number is in a popular format.	The number formats that can trigger the dictionary are: 756.3594.2833.18 756.2487.4093.09 @7.56.7811.8967.63@ The number formats that do not trigger the dictionary are: 756.52--30.6913.27 756.88 85.8748.24
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The AHV number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, identifiant national, insurance number, national identifier, national insurance number, AHV number, AHV-Nummer, Personenidentifikationsnummer, Schweizer Registrierungsnummer, Swiss registration number, or AVS.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 756.3594.2833.18 @7.56.7811.8967.63@ The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 756.52--30.6913.27 756.88 85.8748.24 A756.2487.4093.09A

  Close
• Social Security Numbers (US)

  This dictionary detects Social Security Numbers (SSN) from the United States.

  When you add the Social Security Numbers (US) dictionary to a DLP engine, you must configure the match count. To learn more, see Configuring the Match Count.

  See image.

  

  Close

  The popular format for an SSN is a 9-digit number.

  The following are examples of popular formats:

  • NNN-NN-NNNN
  • NNNNNNNNN
  • NNN NN NNNN

  This dictionary does not use a checksum.

  You can also specify whether to include or exclude specific SSNs. You can configure up to 512 SSNs to include or exclude. To learn more, see Editing Predefined DLP Dictionaries.

  To use this dictionary to detect SSNs, the following guidelines apply:

  These guidelines might not apply to certain use cases. You can adjust your dictionary configurations according to your DLP needs.

  • To catch a small exfiltration of numbers in documents:
    • Set a Confidence Score Threshold of High for the dictionary.
    • When adding a dictionary to an engine, set a low value for the dictionary's match count.
  • To catch a large exfiltration of numbers in spreadsheets or other file types:
    • Set a Confidence Score Threshold of Medium for the dictionary.
    • When adding a dictionary to an engine, set a high value for the dictionary’s match count.
  • In general, configuring a dictionary with a Low Confidence Score Threshold and a low match count value results in too many false positives. Zscaler recommends setting a Confidence Score Threshold of High when its match count value is low.
  • Rich Text Format (RTF) files contain formatting code that can mimic credit card and social security numbers, affecting when a DLP rule is triggered. Plain text files do not contain this formatting code, therefore the DLP rule triggers as expected. So that the DLP policy triggers if confidential numbers are leaked in RTF files, do one of the following steps:
    • Set any value greater than 1 for the dictionary’s match count in the engine.
    • Set a Confidence Score Threshold value of High.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if: The SSN is issued regardless of formatting. You can confirm that the group number is within the range issued for a given area at https://www.ssa.gov/employer/ssnweb.htm. The SSN matches a valid range.	The number formats that can trigger the dictionary are: SA_332831997@@## SA408456050@ 15...21----587---44 ---23527 1165 ...688781663...
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The SSN number is in a popular format.	The number formats that can trigger the dictionary are: SA_332831997@@ SA@408456050@ ...688781663... ??292108209// 269865658 The number formats that do not trigger the dictionary are: 15...21----587---44 ---23527 1165
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The SSN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, date of birth, social security number, or tax payer.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: SA_332831997@@ 408456050 292108209// 269865658 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 15...21----587---44 ---23527 1165

  If a violation is enough to trigger a DLP policy and you have configured a DLP email notification, your auditors receive an email. If you have also included the ${DLPTRIGGERS} macro, the email includes what content triggered the violation.

  Close
• Source Code

  This dictionary detects content related to source code. It detects typical source code and programming phrases such as class implements, class extends, and private void.

  This dictionary does not trigger unless the data size of the detected content is at least 1 KB.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close
• Standardized Bank Code (Mexico)

  This dictionary detects Standardized Bank Code numbers (CLABE) from Mexico.

  The popular format for a Standardized Bank Code number is an 18-character number. The Standardized Bank Code number has the following structure:

  • 3 digits: Bank Code
  • 3 digits: Brach Office Code
  • 11 digits: Account Number
  • 1 digit: Control Digit

  This dictionary uses the Luhn variation checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Standardized Bank Code number matches a valid range. The Standardized Bank Code number can contain: A period, hyphen, or space as delimiters. Multiple periods are allowed. An alphabetical boundary.	The number formats that can trigger the dictionary are: 014027000005555558 00201007--777777-7771 A number format like 032180aaa000118359719 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Standardized Bank Code number is in a popular format. The Standardized Bank Code number must contain a non-alphanumeric boundary. The starting delimiter and ending delimiter can not be an alphabet.	The number formats that can trigger the dictionary are: 014027000005555558 002010077777777771 A number format like A032180aaa000118359719S does not trigger the dictionary.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Standardized Bank Code number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, CLABE, CLABE interbancaria, or standardized bank code.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 014027000005555558 002010077777777771 A number format like @032180aaa000118359719A does not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases.

  Close
• Tax Document

  This dictionary detects tax documents, such as 1040 forms, 1099 forms, 1998-T forms, 3921 forms, etc.

  Zscaler supports only the following document types for tax documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Tax File Numbers (Australia)

  This dictionary detects Tax File Numbers (TFN) from Australia. The TFN is issued by the Australian Taxation Office (ATO) to each taxpaying entity such as an individual, company, superannuation fund, partnership, or trust.

  The popular format for a TFN is a unique 8-digit or 9-digit number.

  This dictionary uses the Mod 11 Check Digit checksum. This checksum is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the TFN matches a valid range. The TFN can contain a period, hyphen, or space as delimiters.	The number formats that can trigger the dictionary are: 371 186 29 459 - 599- .-230 112.474-082 565051603 85655805' '123456782 A number format like 23456782 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The TFN is in a popular format.	The number formats that can trigger the dictionary are: 371 186 29d 459599230 565051603* '123456782 The number formats that do not trigger the dictionary are: 112.474-082 D85655805D
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The TFN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, australian business number, marginal tax rate, medicare levy, portfolio number, service veterans, withholding tax, individual tax return, or tax file number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 371 186 29d 459599230 565051603* '123456782 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 112.474-082 D85655805D

  Close
• Tax Identification Number (Austria)

  This dictionary detects Austrian Tax Identification Numbers (ATTIN), which are 9-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the number matches a valid range.	The number formats that can trigger the dictionary are: 93-173-6581 93/173/6581 93--173/6581 A number format like 931736582 does not trigger the dictionary because it uses an invalid checksum.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The number is in a popular format.	The number formats that can trigger the dictionary are: 931736581 93-173/6581 The number formats that do not trigger the dictionary are: 93-173-6581 93/173/6581 93--173/6581
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The ATTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, österreich, Steuernummer, TIN.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 931736581 93-173/6581 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 93-173-6581 93/173/6581 93--173/6581

  Close
• Tax Identification Number (Belgium)

  This dictionary detects Belgian Tax Identification Numbers (BTIN), which are 11-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 97 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the BTIN matches a valid range.	The number formats that can trigger the dictionary are: 00.0125-111.19 00.01.2511.1.48 0001251111-9 The number formats that do not trigger the dictionary are: 00.01.25-111.18 (invalid checksum) 00.67.25-111.48 (invalid month)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The BTIN is in a popular format.	The number formats that can trigger the dictionary are: 00.01.25-111.19 00012511119 The number formats that do not trigger the dictionary are: 00.0125-111.19 00.01.2511.1.48 0001251111-9
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The BTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, belasting aantal, numéro d'identification fiscale, bnn#.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 00.01.25-111.19 00012511119 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 00.0125-111.19 00.01.2511.1.48 0001251111-9

  Close
• Tax Identification Number (Denmark)

  This dictionary detects Danish Tax Identification Numbers (DTIN), which are 10-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 11 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the DTIN number matches a valid range.	The number formats that can trigger the dictionary are: 010-111-1113 01016011-11 A number format like 010111-1116 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The DTIN number is in a popular format.	The number formats that can trigger the dictionary are: 010111-1113 0101601111 The number formats that do not trigger the dictionary are: 010-111-1113 01016011-11
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The DTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, centrale personregister, civilt registreringssystem, cpr, cpr#, gesundheitskarte nummer, gesundheitsversicherungkarte nummer, TIN, TAX ID, skat id, skattenummer, skat identifikationsnummer, skat kode.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 010-111-1113 01016011-11 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 010111-1113 0101601111

  Close
• Tax Identification Number (Finland)

  This dictionary detects Finland Tax Identification Numbers (FITIN), which are 10-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Modulo 31 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the FITIN number matches a valid range.	The number formats that can trigger the dictionary are: 13 1052-308T 1310 52-308T 13 10 52-308T The number formats that do not trigger the dictionary are: 131052-308U (bad checksum) 131352-3087 (invalid birth month) 321052-3082 (invalid birth date)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The FITIN number is in a popular format.	The number formats that can trigger the dictionary are: 131052-308T 131052+308T 131052a308T 131052A308T The number formats that do not trigger the dictionary are: 13 1052-308T 1310 52-308T 13 10 52-308T
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The FITIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Tunnus Kod, tunnistenumero, tunnus numero, tunnusluku, tunnusnumero, and TIN.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 131052-308T 131052+308T 131052a308T 131052A308T The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 13 1052-308T 1310 52-308T 13 10 52-308T

  Close
• Tax Identification Number (France)

  This dictionary detects French Tax Identification Numbers (FRTIN), which are 13-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 511 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the SPI number matches a valid range.	The number formats that can trigger the dictionary are: 3 023 217 600 053 30--23-217-600-053 30 23-217.600/053 A number format like 3 023 217 600 054 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The SPI number is in a popular format.	The number formats that can trigger the dictionary are: 3023217600053 30 23 217 600 053 30-23-217-600-053 30.23.217.600.053 30/23/217/600/053 The number formats that do not trigger the dictionary are: 3 023 217 600 053 30--23-217-600-053 30 23-217.600/053
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The FRTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, tax identification number and numéro SPI.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 30 23 217 600 053 30-23-217-600-053 30.23.217.600.053 30/23/217/600/053 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 3 023 217 600 053 30--23-217-600-053 30 23-217.600/053

  Close
• Tax Identification Number (Germany)

  This dictionary detects German Tax Identification Numbers (GTIN), which are 11-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 11 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the TIN matches a valid range.	The number formats that can trigger the dictionary are: 65929 970 489 86-095742-719 A number format like 65 929 970 480 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The TIN is in a popular format.	The number formats that can trigger the dictionary are: 65 929 970 489 65929970489 The number formats that do not trigger the dictionary are: 65929 970 489 86-095742-719
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The GTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, identifikationsnummer, steueridentifikationsnummer, steuernummer.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 3023217600053 30 23 217 600 053 30-23-217-600-053 30.23.217.600.053 30/23/217/600/053 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 3 023 217 600 053 30--23-217-600-053 30 23-217.600/053

  Close
• Tax Identification Number (Greece)

  This dictionary detects Greek Tax Identification numbers (GRTIN).

  This dictionary uses a variation of the Modulo 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the GRTIN number matches a valid range.	The number formats that can trigger the dictionary are: 19 86 02 355 19 86-02355 19 8602355 A number format like 198602356 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The GRTIN number is in a popular format.	A number format like 198602355 can trigger the dictionary. The number formats that do not trigger the dictionary are: 19 86 02 355 19 86-02355 19 8602355
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The GRTIN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, afm, afm#, tax identification number, and tin.	A number format like 198602355 can trigger the dictionary. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 19 86 02 355 19 86-02355 19 8602355

  Close
• Tax Identification Number (Hungary)

  This dictionary detects Hungarian Tax Identification Numbers (HUTIN), which are 10-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 97 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the TIN matches a valid range.	The number formats that can trigger the dictionary are: 80 71 592 153 807159215 3 8 071592153 A number format like 8071592154 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The TIN is in a popular format.	A number format like 8071592153 triggers the dictionary. The number formats that do not trigger the dictionary are: 80 71 592 153 807159215 3 8 071592153
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The HUTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, adószám and adóazonosító szám.	A number format like 8071592153 triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 80 71 592 153 807159215 3 8 071592153

  Close
• Tax Identification Number (Indonesia)

  This dictionary detects Tax Identification numbers (NPWP) from Indonesia.

  The popular format for a Tax Identification number is a 15-digit number. The Tax Identification number has the following structure:

  • The first 9 digits (1-9) are the unique identity of the taxpayer.
  • The next 3 digits (10-12) are the code for the tax office where the taxpayer registered for the first time.
  • The last 3 digits (13-15) are the code for central or branch status.

  This dictionary uses the Luhn variation checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	The dictionary counts an instance as a violation if the Tax Identification number matches a valid range.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Tax Identification number is in a popular format.
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Tax Identification number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, npwp, indonesia tax number, indonesian tax number.

  Close
• Tax Identification Number (Ireland)

  This dictionary detects Irish Tax Identification Numbers (IETIN), which are 8- or 9-character unique identifiers issued by the Department of Social Protection. The IETIN is also used by the Revenue Commissioners to identify taxpayers.

  This dictionary uses the Modulo 23 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the IETIN matches a valid range.	The number formats that can trigger the dictionary are: 1234567 T 1234567 TW The number formats that do not trigger the dictionary are: 1234567Y (bad checksum) 123456-7TW (doesn't match regex)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The IETIN is in a popular format.	A number format like 1234567T triggers the dictionary. The number formats that do not trigger the dictionary are: 1234567 T 1234567 TW
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The IETIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Personal Public Service Number, PPS num, Tax Identification Number, TIN, pps number, ppsn, ppsno, uimhir phearsanta seirbhíse poiblí, pps uimh, and Uimhir aitheantais phearsanta.	A number format like 1234567T triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 1234567 T 1234567 TW

  Close
• Tax Identification Number (Luxembourg)

  This dictionary detects Luxembourg Tax Identification Numbers (LUTIN), which are 11- or 13-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Luhn and Mod 11 checksums.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the TIN matches a valid range.	The number formats that can trigger the dictionary are: 189 312 010 5732 1 893120105732 189312010573 2 81 738 480 120 81 738 480.120 81 -738/ .4 80...-120 The number formats that do not trigger the dictionary are: 1893120105732 61432208511 (bad checksum)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The TIN is in a popular format.	The number formats that can trigger the dictionary are: 1893120105732 81.738.480.120 81-738-480-120 81738480120 The number formats that do not trigger the dictionary are: 189 312 010 5732 1 893120105732 189312010573 2 81 738 480 120 81 738 480.120 81 -738/ .4 80...-120 81 738 480.120
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The LUTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, sécurité sociale, zinn nummer, carte sécurité sociale, étain, numéro d'étain, étain non, Numéro d'identification fiscal luxembourgeois, tin, zinn, Luxembourg Tax Identifikatiounsnummer, and steier.	The number formats that trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 1893120105732 81.738.480.120 81-738-480-120 81738480120 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 189 312 010 5732 1 893120105732 189312010573 2 81 738 480 120 81 738 480.120 81 -738/ .4 80...-120 81 738 480.120

  Close
• Tax Identification Number (New Zealand)

  This dictionary detects New Zealand Tax Identification Numbers (NZTIN), which are 8- or 9-digit unique identifiers issued by the Inland Revenue Department (IRD).

  This dictionary uses the Mod 11 algorithm for validation.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NZTIN matches a valid range.	The number formats that can trigger the dictionary are: 136-410132 136-4-10-132 49091 850 A number format like 136-410-134 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NZTIN is in a popular format.	The number formats that can trigger the dictionary are: 49 091 850 136410132 136-410-132 The number formats that do not trigger the dictionary are: 136-410132 136-4-10-132 49091 850
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The NZTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, IRD Number and Inland Revenue Department.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 49 091 850 136410132 136-410-132 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 136-410132 136-4-10-132 49091 850

  Close
• Tax Identification Number (Peru)

  This dictionary detects Tax Identification numbers (PERUC) from Peru.

  This dictionary uses a variation of the Modulo 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the PERUC number matches a valid range.	The number formats that can trigger the dictionary are: 20 503644-968 20--503644-968 2 -05/ .036 449...- 68 A number format like 20503644969 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The PERUC number is in a popular format.	The number formats that can trigger the dictionary are: 20503644968 20 503644 968 The number formats that do not trigger the dictionary are: 20 503644-968 20--503644-968 2 -05/ .036 449...- 68
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PERUC number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Registro Unico Contribuyentes, tax identification number, peru tax identification number, peruvian tax identification number, peru tin, and peruvian tin.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 20503644968 20 503644 968 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 20 503644-968 20--503644-968 2 -05/ .036 449...- 68

  Close
• Tax Identification Number (Poland)

  This dictionary detects Tax Identification numbers (PLTIN) from Poland, which are 10- or 11-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a Member State.

  This dictionary uses the Modulo 11 checksum for the 10-digit number, and the Luhn checksum for the 11-digit number.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the PLTIN number matches a valid range.	The number formats that can trigger the dictionary are: 2234 567895 0207080 3628 A number format like 2234567894 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The PLTIN number is in a popular format.	The number formats that can trigger the dictionary are: 2234567895 02070803628 The number formats that do not trigger the dictionary are: 2234 567895 0207080 3628
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PLTIN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, nip, nip#, numer identyfikacji podatkowej, numeridentyfikacjipodatkowej, tax identification number, tax number, and tin.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 2234567895 02070803628 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 2234 567895 0207080 3628

  Close
• Tax Identification Number (Portugal)

  This dictionary detects Portuguese Tax Identification Numbers (PTIN), which are 9-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Mod 11 algorithm for validation.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the NIF number matches a valid range.	The number formats that can trigger the dictionary are: 299 999998 2544964.40 2544.96350 A number format like 299999991 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The NIF number is in a popular format.	A number format like 299999998 triggers the dictionary. The number formats that do not trigger the dictionary are: 640130.3331 640823 3234 640 8233234
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The PTIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, cpf#, cpf, nif#, nif, numero fiscal, tin.	A number format like 299999998 triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 640130.3331 640823 3234 640 8233234

  Close
• Tax Identification Number (Spain)

  This dictionary detects Tax Identification numbers (CIF) from Spain.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the CIF number matches a valid range.	A number format like A58818501 can trigger the dictionary. The number formats that do not trigger the dictionary are: A 58818501 A588 18501 A588185-01 A58818502 (bad checksum)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The CIF number is in a popular format.	A number format like A58818501 can trigger the dictionary. The number formats that do not trigger the dictionary are: A 58818501 A588 18501 A588185-01
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The CIF number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, tax identification code, Código identificación fiscal, CIF, and CIF número.	A number format like A58818501 can trigger the dictionary. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: A 58818501 A588 18501 A588185-01

  Close
• Tax Identification Number (Sweden)

  This dictionary detects Swedish Tax Identification Numbers (STIN), which are 10-digit unique identifiers issued by the Directorate-General Taxation and Customs Union (DG TAXUD) to all taxpaying individuals of a member state.

  This dictionary uses the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the TIN matches a valid range.	The number formats that can trigger the dictionary are: 640130.3331 640823 3234 640 8233234 A number format like 640823-3235 does not trigger the dictionary.
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The TIN is in a popular format.	The number formats that can trigger the dictionary are: 640130-3331 6408233234 The number formats that do not trigger the dictionary are: 640130.3331 640823 3234 640 8233234
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The STIN is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, personnummer, skatt identifikation, skattebetalarens, identifikationsnummer, sverige tin.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 640130-3331 6408233234 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 640130.3331 640823 3234 640 8233234

  Close
• Tax Identification Number (US)

  This dictionary detects Tax Identification numbers (USITIN) from the United States.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the USITIN number matches a valid range.	The number formats that can trigger the dictionary are: 927 70-5828 927 70 5828 927 70 58 28 A number format like 827705828 does not trigger the dictionary (first digit must be a 9).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The USITIN number is in a popular format.	A number format like 927705828 can trigger the dictionary. The number formats that do not trigger the dictionary are: 927 70-5828 927 70 5828 927 70 58 28
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The USITIN number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, ITIN, USITIN, tax identification number, and taxpayer identification number.	A number format like 927705828 can trigger the dictionary. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 927 70-5828 927 70 5828 927 70 58 28

  Close
• Technical Document

  This dictionary detects technical documents, like computer user manuals, white papers, technical publications, etc.

  Zscaler supports only the following document types for technical documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Transportation and Motor Department Document

  This dictionary detects transportation and motor department documents, like sale or transfer of a vehicle, license forms, driving records, etc.

  Zscaler supports only the following document types for transportation and motor department documents: RTF, PDF, MSG, DOC, DOCX, DOCM, DOTX, DOTM, XLS, XLSX, XLSM, XLTM, PPT, PPTX, PPSX, PPTM, POTM, POTX, and IWORK (pages, numbers, and keynote). The minimum size required for a document is 1 KB.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria
  Low	This dictionary counts an instance as a violation if the Machine Learning (ML) match score is 50 or more.
  Medium	This dictionary counts an instance as a violation if the ML match score is 70 or more.
  High	This dictionary counts an instance as a violation if the ML match score is 90 or more.

  Close
• Treatments Information

  This dictionary uses phrase matching to detect content related to treatments information. It does not use a checksum.

  You can specify an Action to configure how the dictionary evaluates matching treatment names:

  • Count All: The dictionary counts all matches of the treatment name, including identical treatment names, toward the match count.
  • Count Unique: The dictionary counts each unique match of the treatment name toward the match count only once, regardless of how many times the treatment name appears.

  Close
• Unique Master Citizen Number*

  This dictionary detects the 13-digit Unique Master Citizen Number.

  This dictionary uses the Mod 11 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the Unique Master Citizen Number matches a valid range.	The number formats that can trigger the dictionary are: 0407 9894 04655 04.0798-940 4655 04 -07/ .989 404...-655 A number format like 0407989404651 does not trigger the dictionary (bad checksum).
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The Unique Master Citizen Number is in a popular format.	A number format like 0407989404655 triggers the dictionary. The number formats that do not trigger the dictionary are: 0407 9894 04655 04.0798-940 4655 04 -07/ .989 404...-655
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The Unique Master Citizen Number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, JMBG, ЈМБГ, ЕМБГ, Jedinstveni matični broj građana, EMŠO.	A number format like 0407989404655 triggers the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases. The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: 0407 9894 04655 04.0798-940 4655 04 -07/ .989 404...-655

  Close
• VAT Number (Austria)

  This dictionary detects Austrian value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses a variation of the Mod 10 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: AT U10223006 AT-U10223006 AT.U10223006 AT -. U10223006 The number formats that do not trigger the dictionary are: ATU10223007 (invalid checksum) A TU10223006 ATU 10223006 ATU10 223 00 6
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: ATU10223006 AT U10223006 The number formats that do not trigger the dictionary are: AT U10223006 AT-U10223006 AT.U10223006 AT -. U10223006
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by phrases such as: Umsatzsteuer-Identifikationsnummer, Ust-Identifikationsnummer, umsatzsteuer, vat number, vat num, vat.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: ATU10223006 AT U10223006 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: AT U10223006 AT-U10223006 AT.U10223006 AT -. U10223006

  Close
• VAT Number (Belgium)

  This dictionary detects Belgian value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses the Mod 97 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: BE 0776091951 BE-0776091951 BE.0776091951 BE -. 0776091951 The number formats that do not trigger the dictionary are: BE0776091952 (invalid checksum) B E0776091951 BE0 776091951 BE077 609 195 1
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: BE0776091951 BE 0776091951 The number formats that do not trigger the dictionary are: BE 0776091951 BE-0776091951 BE.0776091951 BE -. 0776091951
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, vat number and vat num.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: BE0776091951 BE 0776091951 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: BE 0776091951 BE-0776091951 BE.0776091951 BE -. 0776091951

  Close
• VAT Number (France)

  This dictionary detects French value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses the Mod 97 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: FR 00300076965 FR-00300076965 FR.00300076965 FR -. 00300076965 The number formats that do not trigger the dictionary are: FR00300076964 (invalid checksum) F R00300076965 FR0 0300076965 FR003 000 769 65
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: FR00300076965 FR 00300076965 The number formats that do not trigger the dictionary are: FR 00300076965 FR-00300076965 FR.00300076965 FR -. 00300076965
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Numéro TVA intracommunautaire, TVA intracommunautaire, TVA, and vat.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: FR00300076965 FR 00300076965 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: FR 00300076965 FR-00300076965 FR.00300076965 FR -. 00300076965

  Close
• VAT Number (Germany)

  This dictionary detects German value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses a combination of the Mod 10 and Mod 11 checksums.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: DE 111111125 DE-111111125 DE.111111125 DE -. 111111125 The number formats that do not trigger the dictionary are: D E111111125 DE1 11111125 DE 111 111 125
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: DE111111125 DE 111111125 The number formats that do not trigger the dictionary are: DE 111111125 DE-111111125 DE.111111125 DE -. 111111125
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, mwst, mehrwertsteuer identifikationsnummer, mehrwertsteuer nummer, vat num, vat#.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: DE111111125 DE 111111125 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: DE 111111125 DE-111111125 DE.111111125 DE -. 111111125

  Close
• VAT Number (Ireland)

  This dictionary detects Irish value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses the Modulo 23 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: IE 8Z49289F IE-8Z49289F IE.8Z49289F IE -. 8Z49289F The number formats that do not trigger the dictionary are: IE8Z49289G (bad checksum) I E8Z49289F (doesn't match regex) IE8Z 49289F (doesn't match regex) IE8-Z4-92-89F (doesn't match regex)
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: IE8Z49289F IE 8Z49289F The number formats that do not trigger the dictionary are: IE 8Z49289F IE-8Z49289F IE.8Z49289F IE -. 8Z49289F
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, vat num and vat number.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: IE8Z49289F IE 8Z49289F The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: IE 8Z49289F IE-8Z49289F IE.8Z49289F IE -. 8Z49289F

  Close
• VAT Number (Luxembourg)

  This dictionary detects Luxembourg value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses the Mod 89 checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: LU 13669580 LU-13669580 LU.13669580 LU -. 13669580 The number formats that do not trigger the dictionary are: LU13669581 (invalid checksum) L U13669580 LU1 3669580 LU136 695 80
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: LU13669580 LU 13669580 The number formats that do not trigger the dictionary are: LU 13669580 LU-13669580 LU.13669580 LU -. 13669580
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, tva, vat num, and vat.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: LU13669580 LU 13669580 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: LU 13669580 LU-13669580 LU.13669580 LU -. 13669580

  Close
• VAT Number (Netherlands)

  This dictionary detects Netherlands value-added tax (VAT) numbers, which are alphanumeric identifiers used to identify taxable persons (business) or non-taxable legal entities.

  This dictionary uses the Mod 11 checksum, which is similar to the Luhn checksum.

  The following table lists the confidence score threshold criteria for this dictionary. You can modify the confidence score threshold.

  Confidence Score	Threshold Criteria	Examples of Data
  Low	The dictionary counts an instance as a violation if the VAT number matches a valid range.	The number formats that can trigger the dictionary are: NL 123456782B01 NL-123456782B01 NL.123456782B01 NL -. 123456782B01 The number formats that do not trigger the dictionary are: NL123456783B01 (invalid checksum) N L123456782B01 NL1 23456782B01 NL123 456 78 2B0 1
  Medium	The dictionary counts an instance as a violation if: The requirements of Low Confidence are met. The VAT number is in a popular format.	The number formats that can trigger the dictionary are: NL123456782B01 NL 123456782B01 The number formats that do not trigger the dictionary are: NL 123456782B01 NL-123456782B01 NL.123456782B01 NL -. 123456782B01
  High	The dictionary counts an instance as a violation if: The requirements of Medium Confidence are met. The VAT number is accompanied by any of the dictionary’s default or custom high confidence phrases. For example, Btw-nummer, Btw-num, vat num, and vat.	The number formats that can trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: NL123456782B01 NL 123456782B01 The number formats that do not trigger the dictionary if accompanied by any of the dictionary’s default or custom high confidence phrases are: NL 123456782B01 NL-123456782B01 NL.123456782B01 NL -. 123456782B01

  Close
• Weapons

  This dictionary detects content related to weapons. It detects weapons such as firearms and other military weapons.

  This dictionary does not use a checksum.

  You can modify the Confidence Score Threshold. Confidence scores inform the dictionary how high it must raise the bar, or threshold, for identifying violations and triggering them. To learn more, see Configuring the Confidence Score Threshold.

  Close