Granted Patent B2

US 11,531,779 · App. 15/837,172 · Granted Dec 20, 2022

Inventor: Richard Douglas LeCour (San Jose, CA)

Assignee: Digital Guardian LLC

G06F21/6245G06F16/353G06F16/90344

Abstract

Claims (42)

1. A system for identifying a personal identifier in content, the system comprising:

memory configured to store a plurality of entity definitions corresponding to a plurality of types of personal identifiers, each entity definition specifying a corresponding data format for finding a respective type of personal identifier in content, and a rule, the rule specifying:

(i) a corresponding level of correlation for at least one of a respective geographic or linguistic term in the content with the respective type of personal identifier for the corresponding data format, wherein the corresponding level of correlation comprises a numerical multiplicative weight representative of the correlation between the respective type of personal identifier and the at least one of the respective geographic or linguistic term, and

(ii) a corresponding threshold level at which the corresponding level of correlation is to satisfy for the data format to be classified as the respective type of personal identifier; and

an entity engine executable on one or more processors, the entity engine configured to:

receive first content to identify one or more candidate personal identifiers;

compare a text string in the first content and a first data format specified in a first entity definition of the plurality of entity definitions, the first data format corresponding to a first type of personal identifier from the plurality of types of personal identifiers;

determine a distance between the text string and the first data format;

determine a score based on the distance between the text string and the first data format, wherein the score increases as the distance between the text string and the first data format decreases;

determine that the text string in the first content matches the first data format based on the score;

identify a first level of correlation specified by a first rule of the first entity definition for at least one of a geographic or linguistic term in the first content with the first type of personal identifier;

identify a first threshold level specified by the first rule of the first entity definition corresponding to the first type of personal identifier;

determine that the first level of correlation satisfies the first threshold level specified by the first rule for the first type of personal identifier; and

classify, responsive to the determination and the identification, the text string as the first type of personal identifier for preventing data breach or exfiltration.

2. The system of claim 1 , wherein the entity engine is further configured to increase the score by an amount according to the identified at least one of a geographic or linguistic term in the first content correlated to the first type of personal identifier.

3. The system of claim 1 , wherein the entity engine is further configured to increase the score by an amount according to the level of correlation between the identified at least one of a geographic or linguistic term, and the first type of personal identifier.

4. The system of claim 1 , wherein the score is indicative of a likelihood that the text string actually is a personal identifier of the first type.

5. The system of claim 1 , wherein the geographic term comprises a name or identifier of a geographic region, a name of an entity located in or associated with the geographic region, an address of the geographic region, or a physical address, email address or telephone number of an entity located in or associated with the geographic region, wherein the name or identifier of a geographic region comprises one or more of: a postal code, a name or identifier of a place, state, census area, county, township, town, village, borough, parish, municipality, city, country or continent, or a name or identifier of a geological, cultural, historical or linguistic feature of the geographic region.

6. The system of claim 1 , wherein the linguistic term comprises a word, expression or reference relating to a spoken language, written language, culture, dialect or jargon, wherein the word, expression or reference includes at least one of: a letter, a character, a grapheme, a glyph, a ligature, a numeric digit, or a punctuation mark.

7. The system of claim 1 , wherein the entity engine is further configured to find an identifying or descriptive term for the first type of personal identifier, in the first content, and to classify the text string as the first type of personal identifier responsive to finding the identifying or descriptive term in the first content.

8. The system of claim 7 , wherein the entity engine is further configured to find the identifying or descriptive term located within a predefined level of proximity to the text string in the content.

9. A method for identifying a personal identifier in content, the method comprising:

receiving, by an entity engine executable on one or more processors, first content to identify one or more candidate personal identifiers;

comparing, by the entity engine, a text string in the first content and a first data format specified in a first entity definition of the plurality of entity definitions, the first data format corresponding to a first type of personal identifier from the plurality of types of personal identifiers;

determining, by the entity engine, a distance between the text string and the first data format;

determining, by the entity engine, a score based on the distance between the text string and the first data format, wherein the score increases as the distance between the text string and the first data format decreases;

determining, by the entity engine, that the text string in the first content matches the first data format based on the score, each entity definition specifying a corresponding data format for finding a respective type of personal identifier in content, and a rule, the rule specifying:

(i) a corresponding level of correlation for at least one of a respective geographic or linguistic term in the content with the respective type of personal identifier for the corresponding data format, wherein the corresponding level of correlation comprises a numerical multiplicative weight representative of the correlation between the respective type of personal identifier and the at least one of the respective geographic or linguistic term, and

(ii) a corresponding threshold level at which the corresponding level of correlation is to satisfy for the data format to be classified as the respective type of personal identifier,

wherein the first data format corresponds to a first type of personal identifier from a plurality of types of personal identifiers;

identifying, by the entity engine, a first level of correlation specified by a first rule of the first entity definition for at least one of a geographic or linguistic term in the first content with the first type of personal identifier;

identifying, by the entity engine, a first threshold level specified by the first rule of the first entity definition corresponding to the first type of personal identifier;

determining, by the entity engine, that the first level of correlation satisfies the first threshold level specified by the first rule for the first type of personal identifier; and

classifying, by the entity engine responsive to the determination and the identification, the text string as the first type of personal identifier for preventing data breach or exfiltration.

10. The method of claim 9 , further comprising increasing the score by an amount according to the identified at least one of a geographic or linguistic term in the first content correlated to the first type of personal identifier.

11. The method of claim 9 , further comprising increasing the score by an amount according to the level of correlation between the identified at least one of a geographic or linguistic term, and the first type of personal identifier.

12. The method of claim 9 , wherein the score is indicative of a likelihood that the text string actually is a personal identifier of the first type.

13. The method of claim 9 , wherein the geographic term comprises a name or identifier of a geographic region, a name of an entity located in or associated with the geographic region, an address of the geographic region, or a physical address, email address or telephone number of an entity located in or associated with the geographic region.

14. The method of claim 13 , wherein the name or identifier of a geographic region comprises one or more of: a postal code, a name or identifier of a place, state, census area, county, township, town, village, borough, parish, municipality, city, country or continent, or a name or identifier of a geological, cultural, historical or linguistic feature of the geographic region.

15. The method of claim 9 , wherein the linguistic term comprises a word, expression or reference relating to a spoken language, written language, culture, dialect or jargon, wherein word, expression or reference includes at least one of: a letter, a character, a grapheme, a glyph, a ligature, a numeric digit, or a punctuation mark.

16. The method of claim 9 , further comprising finding, by the entity engine, an identifying or descriptive term for the first type of personal identifier, in the first content, and classifying the text string as the first type of personal identifier responsive to finding the identifying or descriptive term in the first content.

17. The method of claim 9 , further comprising finding the identifying or descriptive term located within a predefined level of proximity to the text string in the content.

Assignments (15)

Continuity (1)