A Super-Dimension Approach in ROLAP Environments

Naydenova, Ina. "A Super-Dimension Approach in ROLAP Environments." Serdica Journal of Computing 5.4 (2011): 309-322. <http://eudml.org/doc/219555>.

@article{Naydenova2011,
abstract = {Often the designer of ROLAP applications follows up with the question “can I create a little joiner table with just the two dimension keys and then connect that table to the fact table?” In a classic dimensional model there are two options - (a) both dimensions are modeled independently or (b) two dimensions are combined into a super-dimension with a single key. The second approach is not widely used in ROLAP environments but it is an important sparsity handling method in MOLAP systems. In ROLAP this design technique can also bring storage and performance benefits, although the model becomes more complicated. The dependency between dimensions is a key factor that the designers have to consider when choosing between the two options. In this paper we present the results of our storage and performance experiments over a real life data cubes in reference to these design approaches. Some conclusions are drawn.},
author = {Naydenova, Ina},
journal = {Serdica Journal of Computing},
keywords = {Combine; Composite; Dimension; Concatenation; OLAP},
language = {eng},
number = {4},
pages = {309-322},
publisher = {Institute of Mathematics and Informatics Bulgarian Academy of Sciences},
title = {A Super-Dimension Approach in ROLAP Environments},
url = {http://eudml.org/doc/219555},
volume = {5},
year = {2011},
}

TY - JOUR
AU - Naydenova, Ina
TI - A Super-Dimension Approach in ROLAP Environments
JO - Serdica Journal of Computing
PY - 2011
PB - Institute of Mathematics and Informatics Bulgarian Academy of Sciences
VL - 5
IS - 4
SP - 309
EP - 322
AB - Often the designer of ROLAP applications follows up with the question “can I create a little joiner table with just the two dimension keys and then connect that table to the fact table?” In a classic dimensional model there are two options - (a) both dimensions are modeled independently or (b) two dimensions are combined into a super-dimension with a single key. The second approach is not widely used in ROLAP environments but it is an important sparsity handling method in MOLAP systems. In ROLAP this design technique can also bring storage and performance benefits, although the model becomes more complicated. The dependency between dimensions is a key factor that the designers have to consider when choosing between the two options. In this paper we present the results of our storage and performance experiments over a real life data cubes in reference to these design approaches. Some conclusions are drawn.
LA - eng
KW - Combine; Composite; Dimension; Concatenation; OLAP
UR - http://eudml.org/doc/219555
ER -