RDF Dataset Compression via Pattern-Coverage Snippets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for automatic analysis of RDF datasets struggle to effectively summarize and reuse their complex structure, lacking a comprehensive representation of both entity descriptions and link patterns, which limits understanding and efficient data handling.
Innovation Solution
The method generates a pattern-coverage snippet by converting edges into nodes, forming an entity-link graph, solving a group Steiner tree problem, and ranking patterns by frequency to create a representative subset that covers essential patterns, ensuring connectivity and compactness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the complete RDF dataset is used for analysis, then comprehensive pattern coverage is achieved, but data size and processing complexity increase significantly
Solution Approach 1:
The patent extracts a representative subset of triples from the complete RDF dataset that covers essential patterns. The system identifies and extracts only the necessary triples needed to represent the dataset's structure and semantics, removing redundant information while preserving pattern coverage.
Solution Approach 2:
The patent segments the RDF dataset into meaningful groups based on entity types, relationships, and patterns. By dividing the dataset into structured segments and selecting representative samples from each segment, the system achieves comprehensive pattern coverage with reduced data size.
2Productivity
If a simplified summary of the RDF dataset is generated, then processing speed improves, but pattern coverage and representativeness deteriorate
Solution Approach 1:
The patent replaces traditional mechanical summarization approaches with an intelligent selection system based on pattern analysis. Instead of simply reducing data through compression or sampling, the system uses automated reasoning to select triples that preserve essential patterns, achieving both speed and accuracy.
Solution Approach 2:
The patent changes the selection criteria from random or uniform sampling to pattern-based selective sampling. By adjusting the selection parameters to prioritize triples that exhibit representative patterns, the system maintains pattern coverage while improving processing efficiency.
3Quantity of substance
If existing summarization methods are applied to RDF datasets, then data size is reduced, but the ability to represent both entity descriptions and link patterns is lost
Solution Approach 1:
The patent creates a universal representation framework that simultaneously captures entity descriptions, link patterns, and their interrelationships. The selected subset of triples serves multiple functions: representing individual entities, their relationships, and the overall dataset structure, achieving comprehensive representation with reduced size.
Solution Approach 2:
The patent combines multiple types of information (entity types, properties, relationships, patterns) into a composite representative subset. By integrating diverse informational elements into a unified structured subset, the system preserves the complex structure of the original dataset while reducing data size.
Data Source
AI summary
Computer-implemented method, apparatus and computer program for automatic analysis of a RDF, Resource Description Framework, dataset. The RDF dataset comprising a set of triples, wherein the RDF dataset is provided as an undirected graph comprising nodes and edges, wherein nodes represent entities and edges represent links between entities.