RDF Data Association via Graph Node Collation and Predicate Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for data association in RDF data sets face challenges in accurately identifying predicates due to score calculation hypotheses that lead to incorrect results, especially when dealing with different kinds of data, resulting in low correctness of data association and inadequate adaptation to data from different groups.
Innovation Solution
A data generation method that involves executing a collation process to acquire reference data, identifying nodes and edges in a graph structure, and associating reference source data with destination data based on node and edge definitions, using SPARQL processing to acquire and output subject and predicate values, and calculating scores for predicate probability to improve data association accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If score calculation hypotheses are used to identify predicates in RDF data sets, then the data association process can be automated, but the correctness of data association deteriorates due to incorrect predicate identification
Solution Approach 1:
The patent introduces an intermediary verification step using domain and range definitions as a mediator between the automated score calculation and the final predicate identification. This intermediary layer checks whether the identified subject-predicate-object triples conform to the ontological constraints, thereby correcting errors introduced by automated scoring without manual intervention for each triple.
2Device complexity
If general score calculation methods are applied to all data, then the processing method remains simple, but the adaptability to data from different groups deteriorates
Solution Approach 1:
The patent applies local quality by making the verification process adaptive to different data groups through domain and range definitions specific to each predicate. Instead of using a uniform verification approach, the system tailors the validation criteria to the characteristics of each data group, allowing the same basic framework to handle diverse data types effectively.
Data Source
AI summary
A method for data generation performed by a computer includes: executing a collation process that includes acquiring reference source data representing a first axis and reference destination data representing a second axis and having a graph structure including a node and an edge, and collating a node of the reference destination data using a value of the acquired reference source data; executing a identification process that includes identifying a second node representing a kind of a first node sharing a first edge between a plurality of collation nodes collated by the collation process, and identifying a third node representing a definition of a domain of the first edge between the first node and the collation nodes; and executing a join process that includes associating the reference source data with the reference destination data based on the second node and the third node.