# US 12,008,623 B2 · Systems and methods for user platform based recommendations > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,008,623 B2 - Document ID: US12008623B2 - Status: Granted - Kind: B2 - Application number: 17/213,395 - Filed: 2021-03-26 - Published: 2022-09-29 - Granted: 2024-06-11 - Earliest claimed: 2021-03-26 - Assignees: Capital One Services, LLC - Applicants: Capital One Services, LLC - Inventors: Jeremy Huang; Michelle Emamdie; Derek Bumpas; Jiaxin Guo; Qiaochu Tang - Web page: https://patentgrep.com/patent/US12008623B2 ## CPC - G06Q 30/0629: {by pre-processing results, e.g. ranking or ordering results} - G06F 16/906: Clustering; Classification - G06F 16/9535: Search customisation based on user profiles and personalisation - G06F 16/9536: Search customisation based on social or collaborative filtering - G06Q 30/0627: {by specifying product or service characteristics, e.g. product dimensions} - G06Q 30/0641: {utilising user interfaces specially adapted for shopping} - G06Q 30/0645: Rental transactions; Leasing transactions - 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) ## Abstract Computer-implemented methods and systems include determining vehicle grades for a user by accessing a plurality of user platforms, identifying user-related content linked to the user via the user platforms, extracting user attributes based on the user-related content, applying weights to vehicle attributes in a vehicle recommendation engine, based on the extracted user attributes, generating the vehicle grades based on the weights, and providing the vehicle grades to the user via a vehicle grading platform. ## Claims ### Claim 1 A computer-implemented method for determining vehicle grades for a user, the method comprising: accessing a plurality of user platforms; identifying user-related content linked to the user via the user platforms; extracting user attributes based on the user-related content by: receiving user images associated with the user-related content; determining one or more image attributes of the user images, the one or more image attributes including content of the user images determined by performing image recognition on the user images; determining context associated with the user images; applying the one or more image attributes of the user images and the context associated with the user images to a machine-learning model, the machine-learning model trained to identify user attributes based on both the one or more image attributes of the user images and the context associated with the user images to output the user attributes; and receiving, from the machine-learning model, the outputted user attributes; applying weights to vehicle attributes in a vehicle grading engine, based on the outputted user attributes; generating the vehicle grades based on the weights; and providing the vehicle grades to the user via a vehicle grading platform. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, further comprising generating a user attribute vector based on the outputted user attributes. ### Claim 3 (depends on 2) The computer-implemented method of claim 2, further comprising identifying a user cluster from a plurality of attribute clusters, wherein the user cluster is most closely related to the user attribute vector when compared to the plurality of attribute clusters. ### Claim 4 (depends on 3) The computer-implemented method of claim 3, further comprising determining the weights for the vehicle attributes based on vehicle acquisitions by other users in the user cluster. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, wherein applying the weights comprises determining affinity levels for one or more of the outputted user attributes and determining a weight value based on the affinity levels. ### Claim 6 (depends on 5) The computer-implemented method of claim 5, wherein determining the affinity level comprises determining at least one of a frequency of engagement, a proportion of engagement, a frequency of content generation, or a proportion of content generation. ### Claim 7 (depends on 1) The computer-implemented method of claim 1, wherein accessing the plurality of user platforms comprises obtaining user permission to access the plurality of user platforms. ### Claim 8 (depends on 7) The computer-implemented method of claim 7, wherein obtaining the user permission comprises requesting the user permission via the vehicle grading platform. ### Claim 9 (depends on 1) The computer-implemented method of claim 1, wherein determining context associated with the user images further comprises: receiving user text associated with the user; and determining content of the user text using a contextual engine. ### Claim 10 (depends on 9) The computer-implemented method of claim 9, further comprising: generating a correlation score between the content of the user text to the content of a plurality of other users; identifying one or more other users who's correlation score is higher than a correlation threshold; and identifying the user attributes further based on attributes of the one or more other users who's correlation score is higher than the correlation threshold. ### Claim 11 (depends on 9) The computer-implemented method of claim 9, further comprising: determining a correlation score between the content of the user text to vehicle reviews for a plurality of vehicles; identifying one or more vehicles with vehicle review correlation scores higher than a correlation threshold; and applying the weights to the vehicle attributes also based on the vehicles with vehicle correlation scores higher than the correlation threshold. ### Claim 12 (depends on 1) The computer-implemented method of claim 1, further comprising updating the weights for the vehicle attributes based on user activity. ### Claim 13 (depends on 1) The computer-implemented method of claim 1, wherein the user-related content comprises associated content related to the user's associates, and further comprising: extracting associated attributes based on the associated content; and applying the weights to vehicle attributes based also on the extracted associate attributes. ### Claim 14 A system, comprising: a memory storing processor readable instructions; and at least one processor configured to access the memory and execute the processor readable instructions, which when executed by the at least one processor, configures the at least one processor to perform a plurality of functions, including functions for: accessing a plurality of user platforms; identifying user-related content linked to the user via the user platforms; extracting user attributes based on the user-related content by: receiving user images associated with the user-related content; determining one or more image attributes of the user images, the one or more image attributes including content of the user images determined by performing image recognition on the user images; determining context associated with the user images; applying the one or more image attributes of the user images and the context associated with the user images to a machine-learning model, the machine-learning model trained to identify user attributes based on both the one or more image attributes of the user images and the context associated with the user images to output the user attributes; and receiving, from the machine-learning model, the outputted user attributes; applying weights to vehicle attributes in a vehicle grading engine, based on the outputted user attributes; generating the vehicle grades based on the weights; and providing the vehicle grades to the user via a vehicle grading platform. ### Claim 15 (depends on 14) The system of claim 14, the plurality of functions further including: generating a user attribute vector based on the outputted user attributes. ### Claim 16 (depends on 15) The system of claim 15, the plurality of functions further including: identifying a user cluster from a plurality of attribute clusters, wherein the user cluster is most closely related to the user attribute vector when compared to the plurality of attribute clusters. ### Claim 17 (depends on 16) The system of claim 16, the plurality of functions further including: determining the weights for the vehicle attributes based on vehicle acquisitions by other users in the user cluster. ### Claim 18 A non-transitory computer readable storage medium storing instructions to control one or more processors to perform operations, including: accessing a plurality of user platforms; identifying user-related content linked to the user via the user platforms; extracting user attributes based on the user-related content by: receiving user images associated with the user-related content; determining one or more image attributes of the user images, the one or more image attributes including content of the user images determined by performing image recognition on the user images; determining context associated with the user images; applying the one or more image attributes of the user images and the context associated with the user images to a machine-learning model, the machine-learning model trained to identify user attributes based on both the one or more image attributes of the user images and the context associated with the user images to output the user attributes; and receiving, from the machine-learning model, the outputted user attributes; applying weights to vehicle attributes in a vehicle grading engine, based on the outputted user attributes; generating the vehicle grades based on the weights; and providing the vehicle grades to the user via a vehicle grading platform. ### Claim 19 (depends on 18) The non-transitory computer readable storage medium of claim 18, wherein determining context associated with the user images further comprises: receiving user text associated with the user; and determining content of the user text using a contextual engine. ### Claim 20 (depends on 18) The non-transitory computer readable storage medium of claim 18 storing further instruction to control the one or more processors to: determining a correlation score between the content of the user text to vehicle reviews for a plurality of vehicles; identifying one or more vehicles with vehicle review correlation scores higher than a correlation threshold; and applying the weights to the vehicle attributes also based on the vehicles with vehicle correlation scores higher than the correlation threshold.