Generative AI Prompt Optimization Through Item Importance Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional technologies using Large Language Models (LLM) for generating prompts lack accuracy in optimizing the input to generative AI, leading to suboptimal generation of information.
Innovation Solution
An information processing apparatus that acquires evaluation information on generated content and estimates the importance of items within prompts, allowing for the selection and optimization of key information to improve prompt generation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional LLM technology is used to generate prompts, then prompt generation is achieved, but the accuracy of prompt optimization is insufficient
Solution Approach 1:
The patent implements a feedback mechanism where evaluation information from generation outcomes is fed back into the system to estimate item importance degrees. This feedback loop enables continuous improvement of prompt optimization accuracy by using actual generation results to refine future prompt constructions.
Solution Approach 2:
The patent replaces conventional LLM-based prompt generation with a hybrid system that uses evaluation information processing and importance degree estimation mechanisms. This substitution allows for more precise control and optimization of prompt elements based on quantifiable importance metrics rather than relying solely on LLM heuristic generation.
2Measurement precision
If more evaluation information is processed to improve prompt accuracy, then optimization accuracy improves, but processing complexity increases
Solution Approach 1:
The patent extracts and processes only the essential evaluation information needed to determine item importance degrees. By selectively extracting relevant evaluation data rather than processing all available information, the system achieves accurate prompt optimization while minimizing processing complexity and computational overhead.
Solution Approach 2:
The patent segments the evaluation information processing into distinct steps: acquiring evaluation information, estimating item importance degrees, and using these estimates for prompt optimization. This segmentation allows each component to be optimized independently and simplifies the overall processing architecture while maintaining high accuracy.
Data Source
AI summary
An information processing apparatus according to the present application includes an acquisition unit and an estimation unit. The acquisition unit acquires evaluation information that is information indicating evaluation on a service using generation information that is information that is generated by using generative AI based on a prompt that includes pieces of information on a plurality of items. The estimation unit estimates degrees of importance of the plurality of items based on the evaluation information that is acquired by the acquisition unit.