Author:
Susanna E. S. Campher
Affiliation:
School of Computer Science and Information Systems, North-West University, Potchefstroom, South-Africa
Keyword(s):
Data Warehousing, Semantic Data Management, Metadata, Dimensional Modeling.
Abstract:
The era of big data has brought on new challenges to data warehousing. Emerging architectural paradigms such as data fabric, data mesh, lakehouse and logical data warehouse are promoted as solutions to big data analytics challenges. However, such hybrid environments, aimed at offering universal data platforms for analytics, have schemas that tend to grow in size and complexity and become more dynamic and decentralized, having a drastic impact on data management. Data integrity, consistency and clear meaning are compromised in large architectures where traditional (relational) database principles do not apply. This paper proposes an investigation into semantic metadata solutions in modern data warehousing from a (logical) dimensional modeling perspective. The primary goal is to determine which metadata and types of semantics are required to support automated dimensionalization as it is assumed to be a good approach to integrate data with different modalities. A secondary goal is findi
ng a suitable model to represent such metadata and semantics for both human and computer interpretability and use. The proposal includes a description of the research problem, an outline of the objectives, the state of the art, the methdology and assumptions, the exepected outcome and current stage of the research.
(More)