A Matrix Conceptual Data Warehouse Model: An Accounting Case Study

Abstract

Pictograms (icons) in a conceptual model help to increase its expressiveness. Pictograms allow analysts representing, e.g., points, lines, regions, time intervals, trajectories, among other elements. Nonetheless, there is a lack of pictograms for linear algebra elements such as vectors, matrices, and their operations. In this paper, we propose pictograms for these elements and incorporate them into a conceptual data warehouse model. We apply them in a matrix accounting case study where vectors and matrices facilitate the consolidation of financial statements. The case study was implemented in Apache Hive. The implementation showed a simpler formulation and a better performance for the analyzed queries compared to the corresponding ones in a conventional data warehouse.