# US 2013/0254883 A1 · Method and system for information leak prevention

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2013/0254883 A1
- Document ID: US20130254883A1
- Status: Published application
- Kind: A1
- Application number: 13/892,346
- Filed: 2013-05-13
- Published: 2013-09-26
- Earliest claimed: 2005-07-14
- Assignees: PortAuthority Technologies Inc.
- Applicants: PortAuthority Technologies Inc.
- Inventors: Lidror TROYANSKY; Assaf Litai; Sharon Bruckner
- Web page: https://patentgrep.com/patent/US20130254883A1

## CPC

- H04L 63/1425: {Traffic logging, e.g. anomaly detection}
- H04L 63/1408: {by monitoring network traffic (monitoring network traffic per se H04L43/00)}
- H04L 63/0245: {Filtering by information in the payload}
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A method for mitigating false positive type errors while applying an information leak prevention policy to identify important information and to prevent outward leakage. A positive criterion is defined for a positive set, and a negative criterion for a negative set of benign traffic. An ambiguity set contains items showing indications for both positive and negative sets. An ambiguity resolution criterion allows ambiguous items to be placed in/removed from the positive set or negative set. Each information item is searched for matches with the positive set. Each item in the positive set is checked for membership in the ambiguity set. The ambiguity resolution criteria are used for each member of the ambiguity set and to remove items from the positive set accordingly. The leak prevention policy is applied for all items remaining in the positive set thus protecting the important information.

## Claims

### Claim 1

A method for mitigating false positive type errors while applying an information leak prevention policy to identify important information that it is desired to protect and to prevent said important information from leaking outwardly of an organization, the method comprising the computer implemented steps of:
defining at least one positive criterion for a positive set, wherein said positive criterion comprises at least one indicator that a corresponding item contains information the distribution of which may be a possible breach of said information leak prevention policy;
defining at least one negative criterion for a negative set, wherein said negative criterion comprises at least one indicator of benign traffic;
establishing an ambiguity set defined by an intersection between said positive set and said negative set, said intersection containing items showing indications for both said positive set and said negative set, such that information items in said intersection, being all of the information items that belong simultaneously to said positive set and to said negative set, enter said ambiguity set;
defining at least one ambiguity resolution criterion for resolving ambiguity of all members of said ambiguity set, thereby to positively place in or to remove said member from said positive set accordingly;
monitoring and analyzing electronic traffic, where each information item in said traffic is searched for matches with said positive set;
checking each item in said positive set for membership in said ambiguity set;
resolving ambiguities of items in said ambiguity set using one of said ambiguity resolution criterion for each member of the ambiguity set and removing items from the positive set accordingly, said resolving being carried out on the items in all said items in said ambiguity set, being said items which are simultaneously members of said positive set and said negative set; and
applying information leak prevention policy for all items remaining in said positive set following said removal of items using ones of said ambiguity resolution criteria, thereby identifying important information that it is desired to protect and to prevent said important information from leaking outwardly of said organization.

### Claim 2 (depends on 1)

A method according to claim 1 wherein said criterion for said positive set comprises Luhn validation.

### Claim 3 (depends on 1)

A method according to claim 1 wherein said criterion for said positive set comprises a checksum validation.

### Claim 4 (depends on 1)

A method according to claim 1 wherein said criterion for said positive set comprises checking for a membership in a pre-defined list.

### Claim 5 (depends on 1)

A method according to claim 1 wherein said criterion for said positive set comprises a match with a regular expression.

### Claim 6 (depends on 1)

A method according to claim 1 wherein said criterion for said positive set comprises a similarity to at least one document.

### Claim 7 (depends on 1)

A method according to claim 1, wherein said resolving said ambiguity comprises searching for terms related to the positive set in proximity to said matches to said positive set.

### Claim 8 (depends on 1)

A method according to claim 1, wherein said resolving said ambiguity comprises searching for terms related to the negative set in proximity to said matches to said positive set.

### Claim 9 (depends on 1)

A method according to claim 1, wherein said resolving said ambiguity comprises applying an additional validation criterion.

### Claim 10 (depends on 1)

A method according to claim 1, wherein said information leak prevention policy comprises applying at least one of the following actions:
i. block the message;
ii. quarantine the message;
iii. encrypt the message;
iv. archive the message;
v. notify an authorized person about the message, and
vi. log the message.

### Claim 11

A method for mitigating false positive type errors while applying information leak prevention policy, the method comprising the computer-implemented steps of:
a. monitoring and analyzing electronic traffic, where each item is validated using a validation filter, said validation filter being configured to assert with a predetermined level of certainty that a non-validated finding belongs to a negative set;
b. inserting a validated item to a validated set and inserting non validated item into a non-validated set;
c. finding for each member of the validated set another item most similar to said member;
d. denoting said member of the validated set a negative if said another most similar item is a member of the non-validated set;
e. denoting said validated item a positive if said most similar item is also a member of the validated set, and
f. applying information leak prevention policy for all items remains in said positive set, thereby mitigating false positive errors through the transduction of information between known and unknown samples.

### Claim 12 (depends on 11)

A method according to claim 11 wherein said validation filter comprises Luhn validation.

### Claim 13 (depends on 11)

A method according to claim 11 wherein said validation filter comprises a checksum validation.

### Claim 14 (depends on 11)

A method according to claim 11 wherein said validation filter comprises checking for a membership in a pre-defined list.

### Claim 15 (depends on 1)

A method according to claim 1 wherein said validation filter comprises a match with a regular expression.

### Claim 16 (depends on 11)

A method according to claim 11, wherein said information leak prevention policy comprises applying at least one of the following actions:
i. block the message;
ii. quarantine the message;
iii. encrypt the message;
iv. archive the message;
v. notify an authorized person about the message, and
vi. log the message.

### Claim 17

A method for mitigating false positive type errors, while applying an information leak prevention policy to messages comprising a plurality of suspected items, through statistical analysis and hypothesis testing, the method comprising the computer-implemented steps of:
a. applying a validation filter to each of said plurality of items within said messages;
b. computing a number or proportion of items expected to pass said validation inform within said message in the case in which said message is benign;
c. counting the number of valid items and comparing it with said expected number or proportion of validated items;
d. defining as a positive case the message wherein said number of validated items is significantly greater than said expected number or proportion, and
e. applying information leak prevention policy for said positive cases.

### Claim 18 (depends on 1)

A method according to claim 1, wherein said positive set comprises social security numbers and wherein said negative set comprises CUISP identifiers that are numbers valid for the set of social security numbers.

### Claim 19 (depends on 1)

A method according to claim 1, further comprising
a) assessing the potential severity of a given possible breach of said information leak prevention policy, to identify certain threats as potentially severe;
b) for threats identified as potentially severe:
i) assigning a default positive label, and
ii) performing at least one test to validate said potential high severity cases as negative.

### Claim 20

A method for maintaining an approximately constant rate of false positive type errors in traffic comprising events, for a given event validation filter, the method comprising:
a) computing an expected fraction of total events in said traffic to pass validation by said event validation filter in a given period;
b) monitoring and analyzing electronic traffic, where each event is validated using said event validation filter;
c) denoting any period wherein a fraction of validated events is significantly greater than said expected fraction, and
d) performing an additional analysis of events validated in said denoted periods.

### Claim 21 (depends on 20)

A method according to claim 20, wherein said thorough analysis comprises a manual analysis.

### Claim 22 (depends on 17)

The method of claim 17, wherein said plurality of items comprises sufficient items to have statistical significance.

### Claim 23

Apparatus for mitigating false positive type errors while applying an information leak prevention policy to identify important information that it is desired to protect and to prevent said important information from leaking outwardly of an organization, the apparatus comprising:
a criterion definer for defining at least one positive criterion for a positive set of items, wherein said positive criterion comprises at least one indicator of said important information and a possible breach of said information leak prevention policy, and further for defining at least one negative criterion for a negative set, wherein said negative criterion comprises at least one indicator of benign traffic;
a disambiguator unit for establishing an ambiguity set in association with an intersection between said positive set and said negative set, such that all of said information items containing indicators for both said positive set and said negative set lie in said intersection and enter said ambiguity set, and for defining at least one ambiguity resolution criterion for resolving said ambiguity;
said apparatus being configured to monitor and analyzing electronic traffic, where each information item in said traffic is searched for matches with said positive set; to check each item in said positive set for membership in said ambiguity set; and to resolve ambiguities using one of said ambiguity resolution criterion for each member of the ambiguity set, said ambiguity resolving thus being carried out on all items which are simultaneously in said positive set and said negative set, and removing items from the positive set accordingly, thereby to allow said information leak prevention policy to be applied for all items remaining in said positive set following said removal of items using ones of said ambiguity resolution criteria, thus mitigating false positive type errors while applying an information leak prevention policy to identify important information that it is desired to protect and to prevent said important information from leaking outwardly of said organization.
