Assessing the understandability of a data warehouse logical model using a decision-tree approach

Abstract

In recent years, Data Warehouses have become essential for researchers and organizations due to their ability to model the near future. For making strategic decisions, the information stored in a data warehouse must be of good quality. The information quality of a data warehouse can be enhanced by determining the data-model quality used for developing the data warehouse. Multidimensional models have been widely accepted as the foundation of data-modeling. Therefore, we believe that improving multi-dimensional model quality could improve the data warehouse quality. Several quality features (such as understandability, simplicity, maintainability, analyzability, coupling, cohesion etc.) have been defined by researchers for measuring the quality of data models. To judge Multidimensional model quality, metrics can be used as they provide a way for evaluating quality in an objective and consistent manner. A few researchers have proposed metrics for the data warehouse multidimensional model. Though, along with the proposal of metrics, they must be validated (theoretically and empirically) to prove their reliability and practical usefulness. This paper focuses on empirical validation of data warehouse logical model metrics using the data-mining machine-learning technique known as decision tree analysis. The results obtained indicate that a few of the proposed metrics are useful in judging the quality of the multidimensional logical model in terms of understandability.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.