Automatic generation of ETL processes from conceptual models

 

Authors
Muñoz, Lilia; Trujillo, Juan; Mazón, Jose Norberto
Format
Article
Status
publishedVersion
Description

Data warehouses (DW) integrate different data sources in order to give a multidimensional view of them to the decision-maker. To this aim, the ETL (Extraction, Transformation and Load) processes are responsible for extracting data from heterogeneous operational data sources, their transformation (conversion, cleaning, standardization, etc.), and its load in the DW. In recent years, several conceptual modeling approaches have been proposed for designing ETL processes. Although these approaches are very useful for documenting ETL processes and supporting the designer tasks, these proposals fail to give mechanisms to carry out an automatic code generation stage. Such a stage should be required to both avoid fails and save development time in the implementation of complex ETL process. Therefore, in this paper we define an approach for the automatic code generation of ETL processes. To this aim, we align the modeling of ETL processes in DW with MDA (Model Driven Architecture) by formally defining a set of QVT (Query, View, Transformation) transformations.
Data warehouses (DW) integrate different data sources in order to give a multidimensional view of them to the decision-maker. To this aim, the ETL (Extraction, Transformation and Load) processes are responsible for extracting data from heterogeneous operational data sources, their transformation (conversion, cleaning, standardization, etc.), and its load in the DW. In recent years, several conceptual modeling approaches have been proposed for designing ETL processes. Although these approaches are very useful for documenting ETL processes and supporting the designer tasks, these proposals fail to give mechanisms to carry out an automatic code generation stage. Such a stage should be required to both avoid fails and save development time in the implementation of complex ETL process. Therefore, in this paper we define an approach for the automatic code generation of ETL processes. To this aim, we align the modeling of ETL processes in DW with MDA (Model Driven Architecture) by formally defining a set of QVT (Query, View, Transformation) transformations.

Publication Year
2018
Language
eng
Topic
Automatic
ETL processes
conceptual models
Automatic
ETL processes
conceptual models
Repository
RI de Documento Digitales de Acceso Abierto de la UTP
Get full text
https://dl.acm.org/citation.cfm?id=1651298
http://ridda2.utp.ac.pa/handle/123456789/4915
Rights
embargoedAccess
License