Information Processing System for Accurate Field-Specific Answers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large language models often provide incorrect answers or struggle to generate accurate responses when questions relate to specific fields, and training these models with field-specific knowledge is challenging.
Innovation Solution
An information processing system that includes a question acquisition module, a question determination module, an information acquisition module, an answer acquisition module, and an answer output module. This system determines whether a question relates to a specific field, acquires relevant information from a database, and uses a language model to generate answers based on this information, improving answer accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a large language model is used to answer questions directly, then the system can provide natural language responses, but the accuracy of answers in specific fields deteriorates
Solution Approach 1:
The patent introduces an intermediary component (field determination module and information acquisition module) between the user's question and the language model. This intermediary retrieves field-specific information from databases before presenting it to the language model, thereby improving answer accuracy in specific domains while maintaining natural language response capabilities
Solution Approach 2:
The system performs preliminary actions by determining the field of the question and acquiring relevant field-specific information before the language model generates its answer. This preliminary retrieval of accurate information from specialized databases ensures that the language model works with verified data, improving reliability without sacrificing ease of operation
2Reliability
If field-specific knowledge is integrated into the language model, then answer accuracy in specific fields improves, but the complexity of training and system architecture increases
Solution Approach 1:
The patent segments the system into distinct functional modules: a field determination module that identifies the domain of the question, an information acquisition module that retrieves field-specific information from specialized databases, and a language model that generates responses. This segmentation allows each component to specialize without requiring the entire system to be retrained, thus improving reliability while managing complexity
Solution Approach 2:
The language model serves multiple functions: it determines the field of the question, acquires field-specific information, and generates natural language responses. This multi-functionality reduces the need for separate specialized models for each field, thereby improving answer accuracy across multiple domains without proportionally increasing training complexity
Data Source
AI summary
Provided is an information processing system configured to: acquire a question based on input from a user; determine whether the question relates to a predetermined field; acquire answer basis information corresponding to the question from a database relating to the predetermined field; request, when it is determined that the question relates to the predetermined field, a for-answer language model to create an answer to the question based on the acquired answer basis information, and acquire an answer from the for-answer language model; and send information answering the question to the user based on the acquired answer.