Node Relevance Search Using Two-Hop Triple Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for extracting useful information from mechanically extracted component data from text documents require reaggregation and a specific schema, making it difficult to analyze relationships that are not explicitly stated and hard to find, especially in non-structured data environments.
Innovation Solution
A method that allows a computer to display relevant nodes connected by two hops, using a triple representation of {subject, predicate, object} to calculate relevance scores and renormalize nodes to zero hops, enabling advanced search and analysis without reaggregation, by adding internal links and calculating scores based on occurrences of triples.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional statistical analysis methods are used to extract useful information from mechanically extracted component data, then information can be extracted through aggregation, but data must be reaggregated into a specific schema layout and input again, increasing complexity and time consumption
Solution Approach 1:
The patent pre-processes and stores component data in a standardized schema layout in advance, creating a pre-organized data structure that can be directly queried without reaggregation. This preliminary organization eliminates the need for repeated data restructuring and input operations when performing statistical analysis.
Solution Approach 2:
The patent introduces an intermediate data storage layer that holds mechanically extracted component data in a standardized format. This intermediate layer acts as a mediator between text analysis and statistical analysis, allowing direct access to processed data without requiring reaggregation into specific schema layouts for each analysis task.
2Loss of information
If conventional statistical analysis methods are used, then information extraction is possible, but a specific schema must be designed for each purpose, increasing device complexity
Solution Approach 1:
The patent creates a universal data storage schema that can serve multiple statistical analysis purposes simultaneously. Instead of designing separate schemas for different analysis tasks, a single standardized schema layout is designed that accommodates various types of component data, allowing the same data structure to be used for different statistical analyses without redesign.
Solution Approach 2:
The patent pre-defines a comprehensive schema layout that anticipates multiple analysis purposes. By designing a flexible, multi-purpose schema in advance, the system eliminates the need to create and switch between different schemas for different analysis tasks, reducing design complexity and enabling universal data access.
3Speed
If simple search methods are used in non-structured data environments, then search operations are fast, but relationships that are not explicitly stated and difficult to find cannot be clarified
Solution Approach 1:
The patent replaces simple mechanical search operations with statistical analysis methods that operate on pre-organized component data. Instead of relying on explicit text matching, the system uses statistical relationships and patterns in the structured data to identify implicit connections, maintaining speed while enhancing relationship detection capability.
Solution Approach 2:
The patent performs preliminary text analysis and component extraction to structure non-structured data before search operations. By pre-processing the data into a structured format with defined schemas, the system enables both fast search operations and the detection of implicit relationships through statistical analysis of the organized data.
Data Source
AI summary
Embodiments of the present invention is a technique of searching for relevant nodes. This technique may include: in response to selection of a first node, displaying, as first relevant nodes, nodes having a first relevance of at least a predetermined value among nodes connected from the first node by two hops; and, in response to selection of at least one of the first relevant nodes, displaying the selected first relevant node as a second node involving the first node. This technique may further include displaying, as second relevant nodes, nodes having a second relevance of at least a predetermined value among nodes connected from the second node by two hops.