Personalized Digital Human Framework Using ML Inference

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current digital human platforms lack the ability to insightfully analyze customer preferences and predict attributes that would be preferred by customers, resulting in a 'one size fits all' approach that fails to consider cultural, linguistic, and custom differences.

Innovation Solution

A framework that utilizes machine learning to infer customer preferences from real-time and historical interactions, employing a deep neural network-based multi-target classification model to dynamically generate personalized digital humans with attributes such as language, accent, and emotion that align with user preferences.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If a one size fits all approach is used for digital human platforms, then device complexity is reduced and ease of manufacture is improved, but adaptability to different customer preferences deteriorates

Engineering Contradiction:

Improveease of creating digital humanVSAvoidadaptability to customer preferences

Core Design Contradiction:

Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The patent segments customer preferences into distinct attributes including race, age, gender, language, accent, and emotion. By dividing the customization process into these discrete segments, the system enables personalized digital humans without requiring complete redesign of the entire platform, thus maintaining ease of manufacture while improving adaptability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically adjusts digital human attributes based on real-time customer preferences and historical data. The platform transitions from static, fixed configurations to dynamic, adaptable configurations that automatically adjust race, age, gender, language, accent, and emotion based on customer interactions and predicted preferences.

Inventive Principle:

Principle #15Dynamics

2Ease of operation

If customer information is collected for marketing purposes only, then data collection is simple and ease of operation is improved, but the ability to understand customer preferences deteriorates

Engineering Contradiction:

Improveease of data collectionVSAvoidloss of customer preference insights

Core Design Contradiction:

Ease of operationVSLoss of information

Solution Approach 1:

The system implements feedback loops where customer interactions with digital humans are continuously monitored and fed back into the machine learning models. This feedback mechanism transforms basic collected data into actionable preference insights, enabling the system to learn and adapt to individual customer preferences while maintaining simple data collection operations.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces machine learning models and preference prediction engines as intermediaries between raw customer data and actionable insights. These intermediary systems process and analyze collected information to extract meaningful preference patterns, preventing loss of valuable customer understanding while keeping data collection operations simple.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If digital human platforms provide APIs for customization, then adaptability to customer preferences is improved, but the system lacks insightful intelligence to predict preferences, worsening measurement precision

Engineering Contradiction:

Improvecustomization capabilityVSAvoidprecision in predicting customer preferences

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system performs preliminary analysis of customer preferences using machine learning models before actual digital human interactions. By predicting preferred attributes in advance based on historical data and customer profiles, the system prepares optimized digital human configurations proactively, improving prediction precision while maintaining customization adaptability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces manual preference analysis and basic API customization with advanced machine learning-based preference prediction systems. This substitution introduces intelligent algorithms that automatically analyze patterns and predict preferences with higher precision, moving beyond simple mechanical customization interfaces.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250045557A1Smart, customer centric, personalized digital human framework

Publication Date: 2025.02.06 DELL PROD LP

AI summary

One example method includes pre-processing a dataset, the dataset including data and/or metadata indicating attributes of a user, and the dataset also includes data and/or metadata that was generated as a result of an interaction between the user and a computing system, after the dataset is pre-processed, providing the dataset as an input to a machine learning model, using the machine learning model to generate, based on the input, respective target variable value predictions for each target in a group of targets, and each of the targets corresponds to a respective attribute of the user, using the target value variable predictions to create, or modify, a digital human that has attributes corresponding to the attributes of the user, and deploying the digital human so that the digital human is available to interact with the user.