One Size Does not Fit All: When to Use Signature-based Pruning to Improve Template Matching for RDF graphs

TL;DR

This paper introduces RDF-h, a flexible framework that selectively applies signature-based pruning for RDF graph template matching, optimizing performance based on dataset and query characteristics.

Contribution

It proposes a novel adaptive framework that evaluates dataset and query features to decide when to use signature-based pruning, improving efficiency in RDF graph matching.

Findings

RDF-h outperforms traditional methods on various datasets

Signature-based pruning effectiveness varies with dataset structure

Adaptive selection improves query performance and scalability

Abstract

Signature-based pruning is broadly accepted as an effective way to improve query performance of graph template matching on general labeled graphs. Most existing techniques which utilize signature-based pruning claim its benefits on all datasets and queries. However, the effectiveness of signature-based pruning varies greatly among different RDF datasets and highly related with their dataset characteristics. We observe that the performance benefits from signature-based pruning depend not only on the size of the RDF graphs, but also the underlying graph structure and the complexity of queries. This motivates us to propose a flexible RDF querying framework, called RDF-h, which selectively utilizes signature-based pruning by evaluating the characteristics of RDF datasets and query templates. Scalability and efficiency of RDF-h is demonstrated in experimental results using both real and synthetic datasets. Keywords: RDF, Graph Template Matching, Signature-based Pruning