Formalization of materialized view problem in ontology-based databases

Abstract

Materialized views are essential for optimizing the performance of traditional databases and data warehouses by accelerating query responses. However, their substantial storage requirements and the impracticality of materializing all possible views raise the problem of selecting which views to persist, a fundamental physical design challenge. This article presents a rigorous formalization of this problem within the context of semantic databases. The methodology employed includes a comprehensive literature review aimed at identifying the variety of se-mantic database representations. This analysis revealed a significant diversity in data models and query languages used. Based on this analysis, a generic formalization framework is pro-posed. This framework enables the expression of various resolution approaches to the materialized view selection problem, taking into account the specificities of semantic databases. It offers broad applicability to any database management system, providing a common language to describe and compare view selection methods.