A methodological approach to incremental view maintenance for optimizing SPARQL queries in Solid

Abstract

This contribution explores innovative strategies for efficient information retrieval in large-scale Solid deployments through the development of incremental view maintenance techniques. Assuming a pivotal role for web agents and aggregators in managing SPARQL queries within decentralized data architectures, we propose the adaptation of established relational database methodologies, specifically the counting algorithm and propagation rules, for the maintenance of materialized views in RDF (Resource Description Framework) data settings. Our research critically assesses the applicability of these algorithms to RDF, addressing the unique challenges posed by SPARQL and linked data's graph nature. We demonstrate the incremental maintainability of aggregation functions like COUNT, SUM, and AVG, while highlighting the limitations for functions such as MIN, MAX, and SAMPLE. By formulating these methodologies using SPARQL algebra, we set the stage for practical implementations that significantly enhance query response times without necessitating full data re-computation. This approach not only underscores the feasibility of applying relational database concepts to linked data but also sets a foundational framework for future research aimed at optimizing data retrieval processes in Solid-based applications and ecosystems at web-scale.