# US 2020/0396258 A1 · Retrospective learning of communication patterns by machine learning models for discovering abnormal behavior

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

- Number: US 2020/0396258 A1
- Document ID: US20200396258A1
- Status: Published application
- Kind: A1
- Application number: 16/927,335
- Filed: 2020-07-13
- Published: 2020-12-17
- Earliest claimed: 2018-12-19
- Assignees: Abnormal Security Corporation
- Applicants: Abnormal Security Corporation
- Inventors: Sanjay Jeyakumar; Jeshua Alexis Bratman; Dmitry Chechik; Abhijit Bagri; Evan James Reiser; Sanny Xiao Yang Liao; Yu Zhou Lee; Carlos Daniel Gasperi; Kevin Lau; Kai Jing Jiang; Su Li Debbie Tan; Jeremy Kao; Cheng-Lin Yeh
- Web page: https://patentgrep.com/patent/US20200396258A1

## CPC

- H04L 63/20: {for managing network security; network security policies in general (filtering policies H04L63/0227)}
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/09: Supervised learning
- G06N 3/098: Distributed learning, e.g. federated learning
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- H04L 63/102: {Entity profiles}
- H04L 63/1416: {Event detection, e.g. attack signature detection}
- H04L 63/1433: {Vulnerability analysis}
- H04L 63/1441: {Countermeasures against malicious traffic (countermeasures against attacks on cryptographic mechanisms H04L9/002)}
- G06N 3/08: Learning methods
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Conventional email filtering services are not suitable for recognizing sophisticated malicious emails, and therefore may allow sophisticated malicious emails to reach inboxes by mistake. Introduced here are threat detection platforms designed to take an integrative approach to detecting security threats. For example, after receiving input indicative of an approval from an individual to access past email received by employees of an enterprise, a threat detection platform can download past emails to build a machine learning (ML) model that understands the norms of communication with internal contacts (e.g., other employees) and/or external contacts (e.g., vendors). By applying the ML model to incoming email, the threat detection platform can identify security threats in real time in a targeted manner.

## Claims

### Claim 1

A computer-implemented method comprising:
establishing, via an application programming interface, a connection with a storage medium that includes a series of past communications received by an employee of an enterprise;
downloading, via the application programming interface, a first portion of the series of past communications corresponding to a first interval of time into a local processing environment;
building a machine learning (ML) model for the employee by providing the first portion of the series of past communications to the ML model as training data;
receiving, via the application programming interface, a communication addressed to the employee; and
establishing whether the communication represents a security risk by applying the ML model to the communication.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, further comprising:
receiving input indicative of an approval from an administrator associated with the enterprise to access the series of past communications;
wherein said establishing is performed in response to receiving the input.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, wherein the series of past communications includes multiple emails that were delivered to the employee.

### Claim 4 (depends on 1)

The computer-implemented method of claim 1, further comprising:
examining each past communication in the first portion of the series of past communications to establish an attribute; and
providing the attributes derived from the first portion of the series of past communications to the ML model as training data.

### Claim 5 (depends on 1)

The computer-implemented method of claim 1, further comprising:
determining, based on an output produced by the ML model, that the communication represents a security risk; and
characterizing the security risk along multiple dimensions.

### Claim 6 (depends on 5)

The computer-implemented method of claim 5, wherein the multiple dimensions include:
an attacked party,
an attack vector,
an impersonated party,
an impersonation strategy, and
an attack goal.

### Claim 7 (depends on 1)

The computer-implemented method of claim 1, wherein the storage medium is a computer server managed by an entity other than the enterprise.

### Claim 8 (depends on 1)

The computer-implemented method of claim 1, wherein the first portion of the series of past communications includes all emails received by the employee during the first interval of time.

### Claim 9 (depends on 1)

The computer-implemented method of claim 1, further comprising:
downloading, via the application programming interface, a second portion of the series of past emails corresponding to a second interval of time that precedes the first interval of time into the local processing environment; and
establishing whether any emails received during the second interval of time represent security risks by applying the ML model to the second portion of the series of past emails.

### Claim 10 (depends on 1)

The computer-implemented method of claim 1, further comprising:
examining the communication to establish multiple attributes;
generating a statistical profile for the communication,
wherein the statistical profile includes a risk score for each pair of attributes included in the multiple attributes, each risk score being based on riskiness of historical communications involving the corresponding pair of attributes.

### Claim 11

A non-transitory computer-readable medium with instructions stored thereon that, when executed by a processor, cause the processor to perform operations comprising:
collecting data related to incoming emails and/or outgoing emails of a customer corresponding to a past interval of time;
generating a communication profile for the customer based on the data;
receiving an incoming email addressed to the customer;
deriving one or more attributes of the incoming email; and
determining whether the incoming email deviates from past email activity by comparing the one or more attributes to the communication profile.

### Claim 12 (depends on 11)

The non-transitory computer-readable medium of claim 11, wherein the customer is an enterprise for which the communication profile is generated.

### Claim 13 (depends on 11)

The non-transitory computer-readable medium of claim 11, wherein the customer is an employee of an enterprise for whom the communication profile is generated.

### Claim 14 (depends on 11)

The non-transitory computer-readable medium of claim 11, wherein said generating comprises:
deriving at least one attribute from each email corresponding to the past interval of time; and
building the communication profile based on the derived attributes.

### Claim 15 (depends on 11)

The non-transitory computer-readable medium of claim 11, the operations further comprising:
providing deviations in the incoming email to a machine learning (ML) model as input; and
determining whether the incoming email is representative of a security risk based on an output produced by the ML model.

### Claim 16 (depends on 15)

The non-transitory computer-readable medium of claim 15, the operations further comprising:
performing a remediation action responsive to determining that the incoming email is representative of a security risk.

### Claim 17 (depends on 11)

The non-transitory computer-readable medium of claim 11, wherein the one or more attributes include a primary attribute and a secondary attribute.

### Claim 18 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein said deriving comprises:
extracting the primary attribute from the incoming email; and
determining the secondary attribute based on the primary attribute and additional information associated with the customer.

### Claim 19

A computer-implemented method comprising:
obtaining first data associated with a first batch of past emails received by employees of an enterprise;
generating a first batch of events that is representative of the first batch of past emails;
obtaining second data associated with a second batch of past emails labeled by one or more administrators,
wherein each past email in the second batch of past emails is associated with a label that specifies risk posed to the enterprise;
generating a second batch of events that is representative of the second batch of past emails; and
storing the first and second batches of events in a database.

### Claim 20 (depends on 19)

The computer-implemented method of claim 19, wherein said generating of the first batch of events comprises:
converting the first data associated with each past email in the first batch of past emails into a predefined schema that defines an event.

### Claim 21 (depends on 20)

The computer-implemented method of claim 20, further comprising:
calculating a risk metric for each past email in the first batch of past emails; and
appending the risk metric calculated for each past email in the first batch of past emails to the corresponding predefined schema.

### Claim 22 (depends on 21)

The computer-implemented method of claim 21, further comprising:
receiving input indicative of a query for events determined not to represent a threat to security of the enterprise;
examining the database to identify events, if any, determined not to represent a threat to security of the enterprise; and
causing display of the identified events on an interface through which the query was submitted.

### Claim 23 (depends on 21)

The computer-implemented method of claim 21, further comprising:
parsing the database to determine whether any past emails in the first batch of past emails are associated with a risk metric that falls beneath a threshold; and
parsing the database to determine whether any past emails in the second batch of past emails are associated with a label that indicates no risk is posed.

### Claim 24 (depends on 19)

The computer-implemented method of claim 19, further comprising:
receiving input indicative of a query for events having a given attribute; and
examining the database to identify events, if any, that have the given attribute.

### Claim 25 (depends on 24)

The computer-implemented method of claim 24, further comprising:
establishing a count of the identified events; and
causing display of the count on an interface through which the query was submitted.

### Claim 26 (depends on 19)

The computer-implemented method of claim 19, further comprising:
acquiring an incoming email addressed to an employee of the enterprise;
parsing the incoming email to identity an attribute of the email;
examining the database to identify events, if any, that have the attribute; and
estimating risk posed by the incoming email based on the identified events.
