# US 12,182,176 B2 · System and method for intelligent synthetic test data generation

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

- Number: US 12,182,176 B2
- Document ID: US12182176B2
- Status: Granted
- Kind: B2
- Application number: 18/191,104
- Filed: 2023-03-28
- Published: 2024-10-03
- Granted: 2024-12-31
- Earliest claimed: 2023-03-28
- Assignees: Accenture Global Solutions Limited
- Applicants: Accenture Global Solutions Limited
- Inventors: Shashank Kundanmal Gandhi; Mihir S. Nanal; Sarang Padmakar Joshi; Pradnya Desai
- Web page: https://patentgrep.com/patent/US12182176B2

## CPC

- G06F 11/3692: {for test results analysis}
- G06F 16/285: {Clustering or classification}
- G06F 11/3684: {for test design, e.g. generating new test cases}
- G06F 11/3688: {for test execution, e.g. scheduling of test suites}
- G06F 11/3698: Environments for analysis, debugging or testing of software
- G06F 18/24147: {Distances to closest patterns, e.g. nearest neighbour classification}
- G06N 20/00: Machine learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

The systems and method described provide a dynamic and automated process for producing test data for cloud migrations. Metadata for transformation designs with attribute classification and relationship data can be used to generate an enriched design-driven data set. The test data can then be used to validate infrastructure provisioned in multi-cloud environments and accommodate most cloud providers. The system can avoid tedious testing cycles that use non-relevant test data (i.e., unrelated to the specific design requirements).

## Claims

### Claim 1

A computer-implemented method for producing enriched test data, the method comprising:
receiving, at a feature extraction module, a first design dataset for a cloud transformation;
extracting, at the feature extraction module, at least a first set of attributes from the first design dataset in a first data file;
receiving, at an attribute classification module, the first data file;
classifying, at the attribute classification module, each attribute in the first data file as falling under at least one attribute type to produce a first classification dataset;
receiving, at an attribute relationship module, the first data file;
identifying, at the attribute relationship module, a transformation relationship for each attribute in the first data file to produce a first relationship dataset;
receiving, at a data engine, input including the first data file, the first classification dataset, the first relationship dataset, and a metadata set associated with the cloud transformation design; and
automatically creating, at the data engine and based on the input, a design-specific test dataset for performing tests on software undergoing a cloud transformation.

### Claim 2 (depends on 1)

The method of claim 1, wherein the classification of each attribute further comprises:
calculating a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
selecting a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
classifying each attribute as an attribute type based on the selected first design.

### Claim 3 (depends on 1)

The method of claim 1, wherein identification of the transformation relationship for each attribute further comprises:
calculating a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
selecting a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
identifying each transformation relationship based on the selected first design.

### Claim 4 (depends on 1)

The method of claim 1, wherein the attribute type includes one of an associative label, transformative label, simple label, and a static label.

### Claim 5 (depends on 1)

The method of claim 1, wherein the transformation relationship is based on one of concatenation, lookup, literal, sequence, arithmetic, subtraction, and addition.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
receiving, at a machine learning model for the attribute classification module, a first recommendation from a reviewer of the design-specific test dataset; and
re-training the machine learning model based on the first recommendation to improve classification accuracy.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
receiving, at a machine learning model for the attribute relationship module, a first recommendation from a reviewer of the design-specific test dataset; and
re-training the machine learning model based on the first recommendation to improve relationship identification accuracy.

### Claim 8

A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to produce enriched test data by performing the following:
receive, at a feature extraction module, a first design dataset for a cloud transformation;
extract, at the feature extraction module, at least a first set of attributes from the first design dataset in a first data file;
receive, at an attribute classification module, the first data file;
classify, at the attribute classification module, each attribute in the first data file as falling under at least one attribute type to produce a first classification dataset;
receive, at an attribute relationship module, the first data file;
identify, at the attribute relationship module, a transformation relationship for each attribute in the first data file to produce a first relationship dataset;
receive, at a data engine, input including the first data file, the first classification dataset, the first relationship dataset, and a metadata set associated with the cloud transformation design; and
automatically create, at the data engine and based on the input, a design-specific test dataset for performing tests on software undergoing a cloud transformation.

### Claim 9 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the instructions further cause the one or more computers to:
calculate a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
select a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
classify each attribute as an attribute type based on the selected first design.

### Claim 10 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the instructions further cause the one or more computers to:
calculate a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
select a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
identify each transformation relationship based on the selected first design.

### Claim 11 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the attribute type includes one of an associative label, transformative label, simple label, and a static label.

### Claim 12 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the transformation relationship is based on one of concatenation, lookup, literal, sequence, arithmetic, subtraction, and addition.

### Claim 13 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the instructions further cause the one or more computers to:
receive, at a machine learning model for the attribute classification module, a first recommendation from a reviewer of the design-specific test dataset; and
re-train the machine learning model based on the first recommendation to improve classification accuracy.

### Claim 14 (depends on 8)

The non-transitory computer-readable medium storing software of claim 8, wherein the instructions further cause the one or more computers to:
receive, at a machine learning model for the attribute relationship module, a first recommendation from a reviewer of the design-specific test dataset; and
re-train the machine learning model based on the first recommendation to improve relationship identification accuracy.

### Claim 15

A system for producing enriched test data comprising one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to:
receive, at a feature extraction module, a first design dataset for a cloud transformation;
extract, at the feature extraction module, at least a first set of attributes from the first design dataset in a first data file;
receive, at an attribute classification module, the first data file;
classify, at the attribute classification module, each attribute in the first data file as falling under at least one attribute type to produce a first classification dataset;
receive, at an attribute relationship module, the first data file;
identify, at the attribute relationship module, a transformation relationship for each attribute in the first data file to produce a first relationship dataset;
receive, at a data engine, input including the first data file, the first classification dataset, the first relationship dataset, and a metadata set associated with the cloud transformation design; and
automatically create, at the data engine and based on the input, a design-specific test dataset for performing tests on software undergoing a cloud transformation.

### Claim 16 (depends on 15)

The system of claim 15, wherein the instructions further cause the one or more computers to:
calculate a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
select a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
classify each attribute as an attribute type based on the selected first design.

### Claim 17 (depends on 15)

The system of claim 15, wherein the instructions further cause the one or more computers to:
calculate a Euclidean Distance (ED) for each design in the first design dataset and obtaining a first ED set;
select a first design in the first design dataset based on the ED for the first design corresponding to the smallest distance in the first ED set; and
identify each transformation relationship based on the selected first design.

### Claim 18 (depends on 15)

The system of claim 15, wherein the attribute type includes one of an associative label, transformative label, simple label, and a static label.

### Claim 19 (depends on 15)

The system of claim 15, wherein the transformation relationship is based on one of concatenation, lookup, literal, sequence, arithmetic, subtraction, and addition.

### Claim 20 (depends on 15)

The system of claim 15, wherein the instructions further cause the one or more computers to:
receive, at a machine learning model for the attribute classification module, a first recommendation from a reviewer of the design-specific test dataset; and
re-train the machine learning model based on the first recommendation to improve classification accuracy.