# US 2026/0064633 A1 · Computer-based platforms/systems/devices/components and/or objects configured for generating values for data structures and methods of use thereof > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2026/0064633 A1 - Document ID: US20260064633A1 - Status: Published application - Kind: A1 - Application number: 18/824,375 - Filed: 2024-09-04 - Published: 2026-03-05 - Earliest claimed: 2024-09-04 - Assignees: Capital One Services, LLC - Applicants: Capital One Services, LLC - Inventors: Saman Baghestani; Daniel Robert Sapp; Roshni Rao; Bryan Parker; Jonathan Ricklis; Nidhi Pandya; James Stephens; Isaac Gonzalez; Bharath Chekuri - Web page: https://patentgrep.com/patent/US20260064633A1 ## CPC - G06F 16/13: File access structures, e.g. distributed indices (arrangements of input from, or output to, record carriers G06F3/06) - G06F 16/116: {Details of conversion of file system types or formats} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Systems and methods of the present disclosure enable a processor to receive at least one proposed file structure and input the at least one proposed file structure into an instant structure generation model that generates unique candidate file structures. The processor may access accepted historical file structures and input the unique candidate file structures and the accepted historical file structures into at least one structure similarity machine learning model to output a subset of unique candidate file structures. The processor may display a list of the subset of unique candidate file structures via a graphical user interface that includes a one-click action element for each unique candidate file structure. ## Claims ### Claim 1 A method comprising: receiving, by at least one processor, at least one proposed file structure, wherein the at least one proposed file structure comprises a plurality of fields, wherein at least one populated field of the plurality of fields is populated with at least one fixed value; inputting, by the at least one processor, the at least one proposed file structure into an instant structure generation model configured to: generate a plurality of candidate value variations by varying each value of at least one other field of the plurality of fields within a threshold value range of each field of the plurality of fields, the at least one other field being different from the at least one populated field; and generate a plurality of unique candidate file structures based on each combination of each value of the at least one other field while holding the at least one populated field as fixed; accessing, by the at least one processor, a plurality of accepted historical file structures, each accepted historical file structure comprising the plurality of fields having a plurality of accepted historical values; inputting, by the at least one processor, the plurality of unique candidate file structures and the plurality of accepted historical file structures into at least one structure similarity machine learning model to output a subset of unique candidate file structures based at least in part on: a similarity of the plurality of candidate value variations of each unique candidate file structure to the plurality of fields having a plurality of accepted historical values of each accepted historical file structure, and trained structure similarity machine learning model parameters; and displaying, by the at least one processor, a list of the subset of unique candidate file structures via a graphical user interface; wherein the graphical user interface comprises a one-click action element for each unique candidate file structure; wherein the one-click action element is configured to, upon a selection of the one-click action element for a particular unique candidate file structure, cause the at least one processor to generate a particular file based at least in part on the particular unique candidate file structure comprising the at least one populated field and the at least one other field of each associated candidate value variation. ### Claim 2 (depends on 1) The method as recited in claim 1, wherein the displaying the list of the subset of unique candidate file structures comprises: determining, by the at least one processor, a similarity metric associated with each unique candidate file structure of the subset of unique candidate file structures based at least in part on the similarity of the plurality of proposed value variations of each unique candidate file structure; ranking, by the at least one processor, the list of the subset of unique candidate file structures in an order from highest to lowest similarity to the accepted historical values of each accepted historical file structure based at least in part on the similarity metric associated with each unique candidate file structure; and displaying, by the at least one processor, the list of the subset of unique candidate file structures in the order. ### Claim 3 (depends on 1) The method as recited in claim 1, wherein the inputting the plurality of unique candidate file structures into the at least one structure similarity machine learning model to output the subset of unique candidate file structures comprises: calculating, by the at least one processor, for each of the unique candidate file structures, by the at least one structure similarity machine learning model, a percentage likelihood of the unique candidate file structure being accepted. ### Claim 4 (depends on 1) The method as recited in claim 1, wherein the at least one structure similarity machine learning model comprises a random forest classifier. ### Claim 5 (depends on 1) The method as recited in claim 1, further comprising: instructing, by the at least one processor, to display the user graphical interface that is configured to enable a manual editing of the particular file. ### Claim 6 (depends on 1) The method as recited in claim 1, wherein the inputting the at least one proposed file structure into the instant structure generation model comprises: receiving, by the at least one processor, a preferred value for a specified field as input; and generating, by the at least one processor using the instant structure generation model, the plurality of proposed value variations by varying each proposed value of the plurality of proposed values based at least in part on the preferred value for the specified field. ### Claim 7 (depends on 1) The method as recited in claim 1, wherein the plurality of fields comprises a plurality of data items identifying a valuation of a vehicle. ### Claim 8 A method comprising: training, by at least one processor, an instant structure generation model to generate a plurality of unique candidate file structures by: configuring the instant structure generation model to receive as input a candidate file structure that comprises a plurality of fields having a plurality of proposed values and return as output the plurality of unique candidate file structures with a plurality of proposed value variations, wherein the plurality of proposed values comprise at least one populated field having at least fixed value, and at least one other field populated with a respective combination of values of each combination of each value of the at least one other field; and inputting as training data, by the at least one processor, a plurality of accepted historical file structures into the instant structure generation model, wherein each accepted historical file structure comprises the plurality of fields having at least one historical value; receiving, by at least one processor, the candidate file structure; providing, by the at least one processor, the candidate file structure to the trained instant structure generation model; and receiving, by the at least one processor, the plurality of unique candidate file structures as output from the trained instant structure generation model. ### Claim 9 (depends on 8) The method as recited in claim 8, wherein the inputting as training data, by the at least one processor, the plurality of accepted historical file structures into the instant structure generation model comprises: pre-processing, by the at least one processor, the plurality of accepted historical file structures into a format expected by the instant structure generation model. ### Claim 10 (depends on 8) The method as recited in claim 8, wherein the configuring the instant structure generation model comprises configuring the instant structure generation model to: generate a plurality of proposed value variations by varying each proposed value of the plurality of proposed values within a threshold value range of each field of the plurality of fields; and generate the plurality of unique candidate file structures based on each combination of the plurality of proposed value variations. ### Claim 11 (depends on 8) The method as recited in claim 8, further comprising inputting, by the at least one processor, the plurality of unique candidate file structures into at least one structure similarity machine learning model to output a subset of unique candidate file structures based at least in part on: a similarity of the plurality of proposed value variations of each unique candidate file structure to the plurality of fields having a plurality of accepted historical values of each accepted historical file structure; and trained structure similarity machine learning model parameters. ### Claim 12 (depends on 8) The method as recited in claim 8, further comprising displaying, by the at least one processor, a list of a subset of unique candidate file structures via a graphical user interface; wherein the graphical user interface comprises a one-click action element for each unique candidate file structure; wherein selection of the one-click action element for a particular unique candidate file structure causes the at least one processor to generate a particular file based at least in part on the particular unique candidate file structure. ### Claim 13 (depends on 8) The method as recited in claim 8, wherein the plurality of fields comprises a valuation of a vehicle. ### Claim 14 A system comprising: at least one processor in communication with at least one computer readable storage medium having software instructions stored thereon, wherein the software instructions, when executed, cause the at least one processor to perform steps to: receive, by at least one processor, at least one proposed file structure, wherein the at least one proposed file structure comprises a plurality of fields, wherein at least one populated field of the plurality of fields is populated with at least one fixed value; input, by the at least one processor, the at least one proposed file structure into an instant structure generation model configured to: generate a plurality of candidate value variations by varying each value of at least one other field of the plurality of fields within a threshold value range of each field of the plurality of fields, the at least one other field being different from the at least one populated field; and generate a plurality of unique candidate file structures based on each combination of each value of the at least one other field while holding the at least one populated field as fixed; access, by the at least one processor, a plurality of accepted historical file structures, each accepted historical file structure comprising the plurality of fields having a plurality of accepted historical values; input, by the at least one processor, the plurality of unique candidate file structures and the plurality of accepted historical file structures into at least one structure similarity machine learning model to output a subset of unique candidate file structures based at least in part on: a similarity of the plurality of candidate value variations of each unique candidate file structure to the plurality of fields having a plurality of accepted historical values of each accepted historical file structure, and trained structure similarity machine learning model parameters; and display, by the at least one processor, a list of the subset of unique candidate file structures via a graphical user interface; wherein the graphical user interface comprises a one-click action element for each unique candidate file structure; wherein the one-click action element is configured to, upon a selection of the one-click action element for a particular unique candidate file structure, cause the at least one processor to generate a particular file based at least in part on the particular unique candidate file structure comprising the at least one populated field and the at least one other field of each associated candidate value variation. ### Claim 15 (depends on 14) The system as recited in claim 14, wherein the displaying the list of the subset of unique candidate file structures comprises: determining, by the at least one processor, a similarity metric associated with each unique candidate file structure of the subset of unique candidate file structures based at least in part on the similarity of the plurality of proposed value variations of each unique candidate file structure; ranking, by the at least one processor, the list of the subset of unique candidate file structures in an order from highest to lowest similarity to the accepted historical values of each accepted historical file structure based at least in part on the similarity metric associated with each unique candidate file structure; and displaying, by the at least one processor, the list of the subset of unique candidate file structures in the order. ### Claim 16 (depends on 14) The system as recited in claim 14, wherein the inputting the plurality of unique candidate file structures into the at least one structure similarity machine learning model to output the subset of unique candidate file structures comprises: calculating, by the at least one processor, for each of the unique candidate file structures, by the at least one structure similarity machine learning model, a percentage likelihood of the unique candidate file structure being accepted. ### Claim 17 (depends on 14) The system as recited in claim 14, wherein the at least one structure similarity machine learning model comprises a random forest classifier. ### Claim 18 (depends on 14) The system as recited in claim 14, further comprising: instructing, by the at least one processor, to display the user graphical interface that is configured to enable a manual editing of the particular file. ### Claim 19 (depends on 14) The system as recited in claim 14, wherein the inputting the at least one proposed file structure into the instant structure generation model comprises: receiving, by the at least one processor, a preferred value for a specified field as input; and generating, by the at least one processor using the instant structure generation model, the plurality of proposed value variations by varying each proposed value of the plurality of proposed values based at least in part on the preferred value for the specified field. ### Claim 20 (depends on 14) The system as recited in claim 14, wherein the plurality of fields comprises a plurality of data items identifying a valuation of a vehicle.