The teaching of Relational On-Line Analytical Processing (ROLAP) in Advanced Database courses.

Abstract

Databases are taught as a core subject at undergraduate and postgraduate levels in computer science degrees. The subject is often divided into conceptual database design and database technologies including physical database design. One of the problems with teaching the topic of ROLAP and Data Warehousing is the difficulty of providing practical exposure to the students. The approach taken in this work is intended to help students apply their knowledge learnt throughout the subject and gain some practical understanding leading to its application and implementation which furthermore provides the basis for analysis and evaluation at the level of critical analysis that would be expected on an advanced level course and at the highest level of learning taxonomies. In our work we have reverse engineered a large Oracle sample dataset and adapted it for use in the teaching of ROLAP. We have also produced our own smaller sample dataset for simplification and understanding of complex queries involving Slicing, Dicing , Pivoting, Rollup and Cube operations. In our examples ROLAP operations and their alternative SQL approaches are performed to provide multi-dimensional results.