Personalized Feature Amount Inference for Prompt-Free Generation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing natural language generation models require prompt engineering skills to achieve desired outputs, which complicates user interaction and convenience.

Innovation Solution

An information processing apparatus infers personalized feature amounts using a personalized feature amount inference unit, leveraging a background information database to associate user attributes and behavior history with input feature amounts, and a personalized feature amount model to generate tailored outputs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If prompt engineering skills are used to adjust prompts until desired results are obtained, then generation accuracy is improved, but user operation complexity increases

Engineering Contradiction:

Improvegeneration accuracyVSAvoiduser operation complexity

Core Design Contradiction:

Manufacturing precisionVSEase of operation

Solution Approach 1:

The system automatically infers personalized feature amounts based on user attributes and behavior history without requiring manual prompt adjustment. The personalized feature amount inference unit performs self-service by autonomously selecting appropriate features and values, eliminating the need for users to engage in prompt engineering while maintaining high generation accuracy.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The personalized feature amount inference unit acts as an intermediary between the user and the generation model. It translates user attributes and behavior history into optimized feature amounts, serving as a mediator that automatically handles the complex prompt engineering task while the user simply provides basic input information.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If manual prompt tuning is performed to search for prompts and output results of other users, then generation accuracy is improved, but time consumption increases

Engineering Contradiction:

Improvegeneration accuracyVSAvoidtime consumption

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The system performs preliminary action by pre-collecting and organizing user attributes and behavior history in the background information database. The personalized feature amount inference unit uses this pre-prepared information to quickly infer optimal feature amounts, eliminating the need for time-consuming manual prompt tuning and searching while maintaining high generation accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system automatically infers personalized feature amounts based on pre-stored user information without requiring manual intervention. This self-service mechanism rapidly generates accurate results by leveraging accumulated user data, significantly reducing the time users would otherwise spend on prompt engineering and result searching.

Inventive Principle:

Principle #25Self-service

3Ease of operation

If personalized feature amount inference is implemented using background information database and inference models, then user convenience is improved, but system complexity increases

Engineering Contradiction:

Improveuser convenienceVSAvoidsystem complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The system segments the complexity into manageable components: a background information database for storing user attributes and behavior history, a personalized feature amount inference unit for processing, and a generation model for output. This segmentation isolates the complex inference logic from the user interface, improving user convenience while organizing system complexity into modular, maintainable units.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The personalized feature amount inference unit serves as an intermediary layer that handles the complex processing of user attributes and behavior history. This mediator shields users from system complexity by automatically managing the inference process, while the modular architecture of the inference unit keeps the overall system complexity manageable and organized.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20250299381A1Information processing apparatus and feature amount inference method

Publication Date: 2025.09.25 KK TOSHIBA

AI summary

An information processing apparatus according to an embodiment includes a personalized feature amount inference unit configured to infer a second prompt token sequence related to a second feature amount on the basis of data related to background information of a person having an attribute close to an attribute of a first user, among data included in a background information database accumulating background information in which attribute information related to an attribute of a person, a feature amount input by the person, and a behavior history related to an action of the person with respect to a product generated on the basis of the feature amount are associated with each other, and a first feature amount input by the first user.