Assertion-Based Document Retrieval via Knowledge Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines are inadequate in finding related documents that support the reasoning or assertions presented in technical and scientific documents, as they rely solely on keyword searches which may not capture the underlying logic or assertions.
Innovation Solution
A method and system that perform document retrieval through assertion analysis by extracting logical structures and entities from documents, creating objects to represent entities, determining and linking assertions to these objects, and structuring the information in a knowledge graph to support queries and document recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If keyword search is used to find related documents, then search speed is improved, but search precision deteriorates because keyword searches cannot capture the underlying logic or assertions in documents
Solution Approach 1:
The patent segments documents into discrete assertion units with logical structures (premises, conclusions, relationships) rather than treating them as continuous text. This segmentation enables the system to index and search specific assertion components independently, maintaining search speed while improving precision by matching assertions rather than just keywords.
Solution Approach 2:
The patent adds a logical structure dimension to document representation by extracting assertions, premises, conclusions, and relationships. This transforms flat keyword-based search into multi-dimensional assertion-based search, allowing queries to match documents based on logical content rather than surface-level keywords, thereby improving search precision without sacrificing speed.
2Measurement precision
If assertion analysis and knowledge graph construction are implemented, then document retrieval precision is improved, but system complexity increases
Solution Approach 1:
The patent performs assertion extraction, logical structure identification, and knowledge graph construction as preliminary processing steps during document ingestion. By pre-structuring documents into assertion units with defined relationships before retrieval operations, the system enables precise querying without requiring complex real-time analysis during search, thus improving retrieval precision while managing system complexity through upfront preparation.
Solution Approach 2:
The patent introduces a knowledge graph as an intermediary layer between raw documents and search queries. This knowledge graph stores extracted assertions, entities, and logical relationships in a structured format that facilitates efficient querying. The intermediary structure simplifies the retrieval process by providing a pre-processed representation that bridges the gap between unstructured documents and structured queries, improving precision without proportionally increasing complexity.
Data Source
AI summary
Document retrieval through assertion analysis on entities and document fragments is disclosed. A document is received. Logical structures and entities are extracted from the document by parsing the document. For an entity in the extracted entities, an object representing the entity is created, an assertion made in the document associated with the entity is determined, and the assertion is linked to the object representing the entity. A logical structure from the extracted logical structures and content of the logical structure containing the assertion are identified and linked to the object representing the entity.