Skip to Main content Skip to Navigation
Conference papers

Graph-Based ETL Processes For Warehousing Statistical Open Data

Alain Berro 1 Imen Megdiche-Bousarsar 2 Olivier Teste 2
1 IRIT-REVA - Real Expression Artificial Life
IRIT - Institut de recherche en informatique de Toulouse
2 IRIT-SIG - Systèmes d’Informations Généralisées
IRIT - Institut de recherche en informatique de Toulouse
Abstract : Warehousing is a promising mean to cross and analyse Statistical Open Data (SOD). But extracting structures, integrating and defining multidimensional schema from several scattered and heterogeneous tables in the SOD are major problems challenging the traditional ETL (Extract-Transform-Load) processes. In this paper, we present a three step ETL processes which rely on RDF graphs to meet all these problems. In the first step, we automatically extract tables structures and values using a table anatomy ontology. This phase converts structurally heterogeneous tables into a unified RDF graph representation. The second step performs a holistic integration of several semantically heterogeneous RDF graphs. The optimal integration is performed through an Integer Linear Program (ILP). In the third step, system interacts with users to incrementally transform the integrated RDF graph into a multidimensional schema.
Complete list of metadata

Cited literature [16 references]  Display  Hide  Download
Contributor : Open Archive Toulouse Archive Ouverte (oatao) <>
Submitted on : Friday, May 13, 2016 - 2:38:07 PM
Last modification on : Friday, June 11, 2021 - 10:49:50 AM
Long-term archiving on: : Tuesday, November 15, 2016 - 6:54:21 PM


Files produced by the author(s)


Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License



Alain Berro, Imen Megdiche-Bousarsar, Olivier Teste. Graph-Based ETL Processes For Warehousing Statistical Open Data. 17th International Conference on Enterprise Information Systems (ICEIS 2015), Apr 2015, Barcelona, Spain. pp.271-278, ⟨10.5220/0005363302710278⟩. ⟨hal-01310725⟩



Record views


Files downloads