RAG Hierarchy Extension for Accurate Entity Assignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual assignment of entities in complex or dynamic hierarchies is time-consuming, prone to human error, lacks transparency, and fails to provide accurate provenance, leading to inefficiencies, data integrity issues, and reduced trust in hierarchical systems.
Innovation Solution
Leverage retrieval-augmented generation (RAG) with a generative AI model to automatically assign new entities by processing user input, generating embedding vectors, and retrieving context from a knowledge base to suggest hierarchical paths while maintaining categorization rules and providing provenance information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual assignment of entities is performed in complex hierarchies, then human judgment and flexibility are applied, but the process becomes time-consuming and prone to human error
Solution Approach 1:
The patent introduces an intermediary system comprising embedding models, vector databases, and retrieval-augmented generation components that mediate between the entity to be assigned and the hierarchical structure. This intermediary automatically processes entity descriptions, retrieves relevant contextual information from the hierarchy, and generates assignment suggestions, eliminating the need for manual human judgment while maintaining high accuracy through multiple validation mechanisms.
Solution Approach 2:
The patent replaces the mechanical manual process of entity assignment with an automated computational system. Instead of human operators manually reviewing and assigning entities, the system uses embedding models to convert entity descriptions into vectors, retrieves relevant hierarchical context through vector database queries, and automatically generates assignment recommendations, thereby eliminating time consumption and human error.
2Loss of information
If manual assignment processes are used, then human operators can make decisions, but transparency and provenance tracking are lacking
Solution Approach 1:
The patent implements comprehensive feedback mechanisms that track and record every step of the automated entity assignment process. The system generates provenance information including the entity being assigned, the retrieved contextual information from the vector database, the generated assignment suggestion, and the final assignment outcome. This feedback loop ensures complete transparency and auditability of the assignment process while maintaining manageable system complexity through structured information flow.
3Stability of the object's composition
If the hierarchy becomes more complex with intricate relationships, then better organization is achieved, but manual management becomes increasingly difficult
Solution Approach 1:
The patent enables the hierarchical structure to serve itself by automatically retrieving relevant contextual information from its own components through the vector database. When a new entity needs to be assigned, the system queries the vector database using the entity's embedding vector and retrieves relevant hierarchical context without external intervention. This self-service mechanism maintains hierarchy integrity while making the assignment process extremely easy to operate.
4Productivity
If manual assignment is performed, then human oversight is provided, but scalability to handle frequent changes is limited
Solution Approach 1:
The patent transforms the entity assignment problem into a parameter-based computational task by converting entity descriptions into embedding vectors and representing hierarchical relationships as structured data in the vector database. This parameter transformation enables the system to process multiple entities simultaneously through batch operations while maintaining consistent assignment criteria through the standardized vector representation and retrieval mechanisms, thereby achieving both high productivity and reliability.
Data Source
AI summary
The present disclosure relates to an automatic assignment of new entities in an existing hierarchical structure by leveraging retrieval-augmented generation (RAG) that improves the performance of a generative artificial intelligence (AI) model by generating a suggested hierarchical structure. For an automatic assignment, a user input may be received that may include a hierarchical structure and a query for adding one or more new entities to the existing hierarchical structure. For each new embedding vector associated with the query for adding a new entity, a retrieval-augmented generation may generate a context from the hierarchical structure based on a relevance to the query, and an additional context from a knowledge base. The retrieval-augmented generation may input a human-readable prompt, combining query with the context, additional context and a system prompt, to the generative AI model such as a large language model for generating the suggested hierarchical structure.