Information Technology Reference
In-Depth Information
Scenario 4
, in this scenario the same tests were accomplished using the proposed
solution. For all possible situations the value of
PrecoLote
was always accurate,
no summarizability or double counting's problems were detected. The results for
“Grand Total” were also correct independently of the dimensional analysis criteria
selected by the end-user. Main drawback: the slight increase in performance is highly
compensated by the advantages and simplicity of the OLAP queries.
The proposed approach allows designers to abstract complexity at a conceptual
level without taking summarizability into account at the initial design. It also avoids
the specification of more complex schema structures (like Bridge table) to deal with
non-strict fact-dimension relationships at different granularities. The proposed solu-
tion besides dealing with roll-up and drill-down incompleteness problems also avoids
the ETL processes to be engaged in complexity.
5
Conclusions
In this paper a short overview to five DW design-driven approaches was presented.
The paper also concluded that data models got from a single design principle are
usually incomplete, which cannot obtain satisfaction and trust of organizations and
business-users simultaneously. A crucial decision for designing multidimensional
models concerns the grain of facts, determined by fact-dimension relationships. To
avoid erroneous results, a multidimensional model must have a consistent granularity,
which means that every measure in the fact must be determined by all dimensions.
The paper presents an approach with which to transform the conceptual multidimen-
sional model into a summarizability-compliant model that avoids erroneous analysis
of data in OLAP tools. Designers do not need to worry with non-strict relations
between facts and dimensions at the conceptual level because when the model is
transformed into the corresponding implementation, the summarizability problem is
correctly addressed.
The impact of the proposed solution on the OOP-DW case study and the expe-
rience we have gathered are encouraging. The option to use a bridge table, despite
being more flexible, was not considered because of two main reasons. First, the re-
quirements for public procurement would originate double-counting's, with implica-
tions to the ETL process complexity; Secondly, to empower designers to abstract
complexity at a conceptual level and enable them to deal with non-strictness at differ-
ent granularities using a flat multidimensional schema that would be understandable
by business-users.
The proposed method was indeed essential to direct us toward a solution that is both
established in the data and oriented to business needs. We also explained how the tech-
nological-driven approach helped us to overcome some limitation of the OLAP tool and
to provide enough expressivity to specify every complex multidimensional structure,
thus providing complete solutions to solve summarizability and double counting prob-
lems. The users' feedback was very positive: the DW schema is effective and compre-
hensive, it satisfies their different application requirements from user querying,
reporting, to multidimensional analysis and management decision.
Search WWH ::
Custom Search