Towards a more straightforward and more expressive metamodel for SDW modeling

Abstract

There are works that propose metamodels for SDW modeling. However, we observe that most of these works defines metamodels that mix concepts of DW modeling with concepts of the OLAP cube modeling. We disagree with this view, because we understand that a DW is essentially a database, which can be analyzed/queried by any data analysis tool. Moreover, we also note other limitations. For example: the most proposed metamodels 1) do not support important techniques for DW modeling and 2) represent the spatiality in a SDW stereotyping the dimensions and fact tables as spatial or hybrid, rather than simply stereotyping the attributes/measures as spatial. Aiming to solve these problems we propose a more straightforward and more expressive metamodel for SDW modeling, named Spatial Data Warehouse Metamodel (SDWM), which describes the constructors and the restrictions needed to model and validate a SDW schema. As a proof of concept, we implemented a CASE tool according to our metamodel and, with this CASE tool, we designed a SDW for the Meteorological Laboratory from the State of Pernambuco, in Brazil.

Comparative Analysis of State-of-the-Art Spatial Data Warehouse Meta-models - Catching the Expressive Power of SDW Schemas!

“…In this paper, we have provided a comparative analysis of the SDWM approach (Del Aguila et al, 2011;Cuzzocrea & Fidalgo, 2012a;Cuzzocrea & Fidalgo, 2012b) against the state-of-the-art SDW meta-model proposals (Fidalgo et al, 2004;Malinowski & Zimányi, 2007;Glorio & Trujillo, 2008). Results of our analysis clearly state that the SDWM proposal exposes a higher expressive power and allows us to obtain more concise and compact SDW schemas.…”

“…Spatial Data Warehouse Metamodel (SDWM) (Del Aguila et al, 2011;Cuzzocrea & Fidalgo, 2012a;Cuzzocrea & Fidalgo, 2012b) has been provided recently with the goal of effectively supporting the modeling of Spatial Data Warehouses (SDW) (Bédard et al, 2001;Zghal et al, 2003) by adding several contributions. Among these, relevant ones concern with separating the SDW conceptual modeling from the OLAP data cube (Gray et al, 1997) conceptual modeling, supporting SDW complex constructs modeling, and, finally, stereotyping attributes and measures as spatial objects directly.…”

Enhancing Coverage and Expressive Power of Spatial Data Warehousing Modeling: The SDWM Approach

“…degenerate dimensions, role-playing dimensions and bridge tables) and (ii) represents the spatiality in a SDW stereotyping the dimensions and fact tables as spatial or hybrid, rather than directly stereotyping the attributes/measures as spatial. With aim of overcoming such limitations, we have proposed the Spatial Data Warehouse Metamodel (SDWM) [5]. However, due to space restriction, a more elaborated specification of SDWM cannot be made in our previous work.…”