Vehicle Recommendation System Using Machine Learning Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Vehicle buyers face challenges in making informed decisions due to overwhelming options and complex criteria, often relying on biased recommendations from salespersons or limited criteria, leading to purchases that may not meet their needs and preferences.

Innovation Solution

A vehicle recommendation system utilizing a machine learning model that analyzes user data, interactions, and vehicle information to provide personalized recommendations based on user profiles, criteria scores, and vehicle characteristics, while maintaining user privacy and data security.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If customers rely on recommendations from friends or family, then they may get useful guidance, but the recommendations may not be suitable for their specific needs and preferences due to different individual preferences

Engineering Contradiction:

Improveloss of personalized vehicle recommendation informationVSAvoidadaptability of recommendations to individual preferences

Core Design Contradiction:

Loss of informationVSAdaptability or versatility

Solution Approach 1:

The patent introduces a machine learning-based recommendation system as an intermediary between vehicle data and customers. This system processes customer preferences, vehicle characteristics, and transaction data to generate personalized recommendations, serving as an unbiased mediator that adapts to individual needs without being influenced by sales bias or limited personal recommendations

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system dynamically adjusts recommendation parameters based on customer profiles, vehicle attributes, and transaction data. By changing the parameters used in recommendation generation based on individual customer characteristics, the system provides personalized recommendations that adapt to each customer's specific needs and preferences

Inventive Principle:

Principle #35Parameter changes

2Loss of information

If customers rely on information from skilled sales persons or advertising campaigns, then they may receive comprehensive vehicle information, but the information may be biased toward specific makes or models with the intent to make a sale

Engineering Contradiction:

Improveloss of unbiased vehicle informationVSAvoidbias in vehicle recommendations

Core Design Contradiction:

Loss of informationVSObject-generated harmful factors

Solution Approach 1:

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables customers to receive objective, data-driven recommendations without being influenced by sales personnel bias. By using automated machine learning models that process transaction data and vehicle characteristics independently, the system provides self-service recommendations that are not motivated by sales targets

Inventive Principle:

Principle #25Self-service

3Device complexity

If customers simplify the purchase by relying on a limited set of vehicle criteria such as brand and model, then they may reduce decision complexity, but they may not fully capture the complex trade-offs that customers face when making a vehicle purchase

Engineering Contradiction:

Improvecomplexity of vehicle selection processVSAvoidloss of comprehensive vehicle evaluation information

Core Design Contradiction:

Device complexityVSLoss of information

Solution Approach 1:

The patent segments the complex vehicle selection process into manageable components: customer preference profiling, vehicle characteristic analysis, and recommendation generation. By breaking down the decision-making process into these segments, the system reduces perceived complexity while maintaining comprehensive evaluation of multiple criteria including price, performance, fuel efficiency, and safety

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The recommendation system acts as an intermediary that handles the complex analysis of multiple vehicle criteria. It processes comprehensive vehicle data and customer preferences, then presents simplified personalized recommendations, allowing customers to benefit from thorough evaluation without bearing the cognitive burden of analyzing all trade-offs themselves

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20240394773A1Systems and methods for vehicle recommendations

Publication Date: 2024.11.28 CAPITAL ONE SERVICES LLC

AI summary

Systems and method for generating vehicle recommendations are determined using interaction information. An information data set may be mapped to a first user and include interactions based on a period of time. The interactions may be parsed by a trained machine learning model to determine trends and attributes. A user profile and a user score for different criteria may be compared to vehicle sores of multiple vehicles. Comparing the scores may identify recommended vehicles for the first user which may be transmitted to a user device of the user.