# US 2019/0318203 A1 · System and method for custom security predictive methods

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

- Number: US 2019/0318203 A1
- Document ID: US20190318203A1
- Status: Published application
- Kind: A1
- Application number: 16/385,128
- Filed: 2019-04-16
- Published: 2019-10-17
- Earliest claimed: 2018-04-16
- Assignees: Interset Software Inc.
- Applicants: Interset Software, Inc.
- Inventors: Stephan Jou; Mario Daigle; Shaun Pilkington; Michael John Cyze; Josh Mahonin; Wesley Lawrence
- Web page: https://patentgrep.com/patent/US20190318203A1

## CPC

- G06Q 10/0635: Risk analysis of enterprise or organisation activities
- G06F 3/0482: Interaction with lists of selectable items, e.g. menus
- G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade}
- G06F 18/2321: using statistics or function optimisation, e.g. modelling of probability density functions
- G06F 21/566: {Dynamic detection, i.e. detection performed at run-time, e.g. emulation, suspicious activities}
- G06F 21/577: {Assessing vulnerabilities and evaluating computer system security}
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A system and method is described for providing custom predictive models for detecting electronic security threats within an enterprise computer network. The custom models may be defined in a declarative language. The custom models, along with native models, may be combined together to provide custom machine learning (ML) use cases.

## Claims

### Claim 1

A method of processing a custom predictive security model comprising:
retrieving a custom predictive security model definition defining:
input data from one or more available data sources providing security related information, the input data used in the predictive security model; and
model logic providing a scoring function used to compute a predicted output value from the input data;
ingesting, from the available data sources, the input data specified in the retrieved custom predictive security model;
loading the ingested input data into the scoring function of the custom predictive security model; and
outputting a predicted value from the scoring function based on the ingested input data.

### Claim 2 (depends on 1)

The method of claim 1, wherein the custom predictive security model definition further defines one or more data aggregations to be applied to one or more input data, the method further comprising aggregating the input data according to the one or more data aggregations during ingestion.

### Claim 3 (depends on 1)

The method of claim 1, further comprising processing one or more native models each of the native models providing a respective output predictive value based on ingested input data.

### Claim 4 (depends on 3)

The method of claim 3, further comprising processing one or more supermodels, each of the supermodels specifying two or more models and one or more Boolean operators joining the two or more models to provide a predictive value for the supermodel.

### Claim 5 (depends on 4)

The method of claim 4, wherein one or more of the supermodels further define one or more trigger conditions for triggering processing of the respective supermodel.

### Claim 6 (depends on 5)

The method of claim 5, further comprising providing a user interface for creating the one or more supermodels comprising:
trigger selection functionality for selecting one or more triggering events from available triggering events;
model selection functionality for selecting one or more of available native models, custom models and supermodels;
Boolean operator definition functionality for defining Boolean operators joining selected models; and
output definition functionality for defining an output of the supermodel.

### Claim 7 (depends on 6)

The method of claim 6, further comprising:
creating one or more supermodels using the user interface for creating the one or more supermodels; and
storing the created one or more supermodels.

### Claim 8 (depends on 1)

The method of claim 1, further comprising providing a user interface for creating the one or more custom models, the user interface comprising:
import functionality for importing a data schema; and
import functionality for importing model logic.

### Claim 9 (depends on 8)

The method of claim 8, wherein the import functionality for importing model logic imports model logic is defined using Predictive Model Markup Language (PMML).

### Claim 10 (depends on 9)

The method of claim 9, further comprising:
creating one or more custom models using the user interface for creating the one or more custom models; and
storing the created one or more custom models.

### Claim 11

A computing system for processing a custom predictive security model comprising:
a processor for executing instructions;
a memory storing instructions, which when executed by the processor configure the computing system to:
retrieve a custom predictive security model definition defining:
input data from one or more available data sources providing security related information, the input data used in the predictive security mode; and
model logic providing a scoring function used to compute a predicted output value from the input data;
ingest, from the available data sources, the input data specified in the retrieved custom predictive security model;
load the ingested input data into the scoring function of the custom predictive security model; and
output a predicted value from the scoring function based on the ingested input data.

### Claim 12 (depends on 11)

The computing system of claim 11, wherein the custom predictive security model definition further defines one or more data aggregations to be applied to one or more input data, and wherein the instructions stored in memory, when executed by the processor, further configure the computing system to aggregate the input data according to the one or more data aggregations during ingestion.

### Claim 13 (depends on 11)

The computing system of claim 11, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to process one or more native models each of the native models providing a respective output predictive value based on ingested input data.

### Claim 14 (depends on 13)

The computing system of claim 13, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to process one or more supermodels, each of the supermodels specifying two or more models and one or more Boolean operators joining the two or more models to provide a predictive value for the supermodel.

### Claim 15 (depends on 14)

The computing system of claim 14, wherein one or more of the supermodels further define one or more trigger conditions for triggering processing of the respective supermodel.

### Claim 16 (depends on 15)

The computing system of claim 15, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to provide a user interface for creating the one or more supermodels comprising:
trigger selection functionality for selecting one or more triggering events from available triggering events;
model selection functionality for selecting one or more of available native models, custom models and supermodels;
Boolean operator definition functionality for defining Boolean operators joining selected models; and
output definition functionality for defining an output of the supermodel.

### Claim 17 (depends on 16)

The computing system of claim 16, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to:
create one or more supermodels using the user interface for creating the one or more supermodels; and
store the created one or more supermodels.

### Claim 18 (depends on 11)

The computing system of claim 11, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to provide a user interface for creating the one or more custom models, the user interface comprising:
import functionality for importing a data schema; and
import functionality for importing model logic.

### Claim 19 (depends on 18)

The computing system of claim 18, wherein the import functionality for importing model logic imports model logic is defined using Predictive Model Markup Language (PMML).

### Claim 20 (depends on 19)

The computing system of claim 19, wherein the instructions stored in memory, when executed by the processor, further configure the computing system to:
create one or more custom models using the user interface for creating the one or more custom models; and
store the created one or more custom models.

### Claim 21

A non-transitory computer readable memory, storing instructions, which when executed by a processor of a computing system, configure the computing system to:
retrieve a custom predictive security model definition defining:
input data from one or more available data sources providing security related information, the input data used in the predictive security mode; and
model logic providing a scoring function used to compute a predicted output value from the input data;
ingest, from the available data sources, the input data specified in the retrieved custom predictive security model;
load the ingested input data into the scoring function of the custom predictive security model; and
output a predicted value from the scoring function based on the ingested input data.
