Information Processing Device for Secure LLM Query Handling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing answer systems using large-language models (LLMs) face security concerns when handling confidential information, as they require transmitting unlearned data to external servers, potentially leading to information leakage and security risks.

Innovation Solution

An information processing device that locally processes user queries and generates answers using a machine learning model provided by a server device, where the device segments documents into chunks and transmits only relevant, confidential chunks to the server for processing, ensuring secure handling and minimizing data exposure.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If data is transmitted to external server to be provided to LLM, then answer capability for unlearned knowledge is improved, but security risk increases

Engineering Contradiction:

Improveanswer capabilityVSAvoidsecurity risk

Core Design Contradiction:

Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The patent segments documents into smaller units called chunks before transmitting to the server. This segmentation ensures that only necessary portions of data are transmitted, reducing the exposure of confidential information while maintaining the ability to provide accurate answers to unlearned knowledge queries

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent extracts only the relevant chunks of data needed to answer specific queries from the entire document database. This extraction approach minimizes the amount of data transmitted to external servers, thereby reducing security risks while preserving the LLM's capability to handle unlearned knowledge

Inventive Principle:

Principle #2Taking out (Extraction)

2Measurement precision

If entire document is transmitted to server, then answer accuracy is improved, but data exposure increases

Engineering Contradiction:

Improveanswer accuracyVSAvoiddata exposure

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent divides documents into manageable chunks that contain only the relevant information needed for answering queries. This segmentation allows the system to transmit minimal necessary data to the server, reducing data exposure while maintaining answer accuracy through targeted information retrieval

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by transmitting only the necessary portions of documents (chunks) rather than entire documents. This partial transmission approach reduces data exposure while providing sufficient information for accurate answers, avoiding the excessive action of transmitting complete documents

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20250103627A1Information processing device, and information processing method

Publication Date: 2025.03.27 AKUSERU KK

AI summary

There is provided an information processing device for obtaining an answer to a query by processing a document and the query using a machine learning model provided by a server device. The information processing device includes: a reception portion configured to receive the query input by a user using an input device; a search portion configured to search for one or more documents stored in a document database private to the server device based on the received query and configured to acquire one or more contents having high relevance to the query as a search result; a transmission portion configured to transmit the search result to the server device via a communication device; and a receiving portion configured to receive, from the server device via the communication device, an answer sentence generated by the server device using the transmitted search result.